Anatoly M. Khazanov: Nomads and the Outside World

N o m a d s a n d the Outside World

Nomads and the Outside World Second Edition ANATOLY M. KHAZANOV

Translated by

JULIA CROOKENDEN With a Foreword by

ERNEST GELLNER

The University of Wisconsin Press

The University of Wisconsin Press 114 North Murray Street Madison, Wisconsin 53715 3 Henrietta Street London W C 2 E 8LU, England Original Russian version copyright © 1983 V A A P English translation copyright © 1983 Anatoly M. Khazanov Foreword copyright © 1983 Ernest Gellner Introduction to the second edition and Bibliography to the second edition Copyright © 1994 The Board of Regents of the University of Wisconsin System All rights reserved 10

9

8

7

6

5

4

3

2

1

Printed in the United States of America First published in 1984 by Cambridge University Press

Library of Congress Cataloging-in-Publication Data Khazanov, Anatoly M. (Anatoli Michailovich), 1 9 3 7 Nomads and the outside world / A . M . Khazanov; translated by Julia Crookenden; with a foreword by Ernest Gellner.—2nd ed. 442 p. cm. Originally published: Cambridge; N e w York: Cambridge University Press, 1984. With new introd. and bibliography. Includes bibliographical references and index. ISBN 0-299-14284-1 1. Nomads. 2. Transhumance. I. Title. GN387.K4513 1994 305.8 dc20 93-41133

To my wife Irina and my son Yaakov, with love.

Contents

Foreword by Ernest Gellner Acknowledgments Introduction to the second edition

ix xxvii xxix

Introduction The phenomenon of nomadism: myths and problems

1

Nomadism as a distinct form of food-producing economy What is nomadism? The basic forms of pastoralism The species-composition of herds Size of herds Nature of the utilization of ecological zones Character of pastoral migrations Nature of the utilization of the products of pastoral economies and dietary systems The main types of pastoral nomadism Problems of balance and the non-autarky of the pastoral economy

15 15 17 25 28 33 36 38 39 68

2

The origins of pastoral nomadism

85

3

The social preconditions of the relations between nomads and the outside world Native model, theoretical model and reality The problem of ownership in nomadic societies Family, household and community in nomadic societies Kinship and descent in nomadic societies Segmentary systems in nomadic societies The upper levels of sociopolitical organization in nomadic societies Property-inequality and social differentiation in nomadic societies Nomadic chiefdoms The theme and its variations

1

vii

119 119 123 126 138 144 148 152 164 169

Contents 4

5

Modes of nomadic adaptation to the outside world Sedentarization Trade and mediatory trade Submission and the different forms of the dependence of nomads on sedentary societies Subjugation and the different forms of the dependence of sedentary societies on nomads

198 198 202 212 222

Nomads and the state Nomadic statehood and the conditions of its emergence Main types and tendencies of the emergence and evolution of nomadic statehood The Eurasian steppes, semi-deserts and deserts The Middle East The Near East East Africa Conclusions

231 233 263 274 290 295

By way of a conclusion: the outside world and nomads

303

Abbreviations Bibliography Works in languages other than Russian Works in Russian Bibliography to the second edition Works in western languages Works in Russian Index

305 307 307 331 356 356 369 372

viii

228 228

Foreword

A. M. Khazanov's global and comparative study of pastoral nomadism is a quite outstanding scholarly achievement. Astonishingly, this task had not been attempted before. It very much needed doing, and it is a remarkable piece of good fortune that, the first time it was attempted, it should be done so brilliantly. But it is perhaps no accident that it should have been done by a man of Khazanov's background and qualifications. His principal previous work was a social history of the Scythians, and he is, amongst other things, an historian thoroughly at home in the classical sources relevant to our knowledge of the Scythians, the first properly documented pastoral nomads. His intimacy with this classical material is complemented, on the one hand, by a mastery of the documentation relevant to the medieval and modern nomads of the Eurasian steppe, and on the other, by a remarkable familiarity with Western anthropological work on nomads in other parts of the world. The synoptic view which we are offered would not have been possible without this unusual, and perhaps unique, scholarly equipment. But it is not, of course, simply a matter of a mastery of the relevant documentation. The study of pastoral nomads, as of any other subject, only comes to life if inspired by some important question, some haunting problem. Khazanov's study is the fruit not merely of his remarkable range of erudition, but equally of the questions which inevitably haunt the Russian and Soviet intellectual tradition of which he is part. It is of course not surprising that a major study of nomadism should emerge from Russia. Russian history and consciousness - whether through education or through a genuine folk memory - is pervaded by an awareness of the nomad problem, more so presumably than any other European nation. The Magyars may look back romantically to a nomadic past, and their populist ethnographers may seek the nomad origins of institutions still found in nineteenth-century Hungarian villages; but the Russian connection with nomads is deeper and more persistent. The first Russian state was destroyed by nomads; the Muscovite state began its career as fiscal agent of a 1

1

A . M. Khazanov: Sotsiatnaia ix

istariia skifov.

Naukn, Moscow, 1975.

Foreword nomad empire, and when the balance of power tilted away from the inhabitants of the steppe to those of the forest, this tax-collector state first ceased to pay up, keeping its revenue for itself, and then expanded to become, in turn, an empire eventually incorporating and administering a number of nomad societies, including its own erstwhile rulers. The expropriators were in turn expropriated. If the conquest by nomads has left its mark on the Russian soul, mythology and literature, then the conquest of nomads has left behind rich administrative and other records of the functioning of nomad society. It is not merely the Russian background which is relevant. The specific questions asked spring from the sociology of Marxism, and indeed from the wider tradition which has dominated Russian social thought since the nineteenth century, and of which Marxism is but an element. The central theme in that tradition is the notion of progress or social evolution. This was a natural idea in a nineteenth-century milieu: it was the perception of massive and apparently persistent change which inspired social thought and presented the problem. The meaning, mechanisms, direction of that change, if located, would have constituted its solution. In Western anthropology and sociology, a rival theory or approach, known as Functionalism, was to be found. This took social cohesion and persistence as the central datum, and strove to explain it. The best known criticism of that school, so well known as to become a hackneyed and almost a joke phrase, was that it 'failed to account for social change'. The critics of Functionalism sometimes went further and suggested that functionalists had a political motive for ignoring or denying change: perhaps they were paid by the established order to try and stop it, or to discourage anyone from endeavouring to initiate it, by pretending it did not or could not occur, and ought not really to occur; that society was essentially a self-maintaining, self-reproducing system, and that any deviation from this norm was pathological. Functionalism has often been attacked in this manner. Marxism, much criticised in other ways, has strangely seldom been attacked by means of the mirror-image criticism to which it seems conspicuously open: can it account for stagnation? Pastoral nomadism is not the only area in which this question is pertinent. In fact, the most proper way to put the question is in a generic form. Was nineteenth-century thought justified in being so smitten by the fact of change and development as to make it the central trait of social life? Is not change, and in particular sustained cumulative change on the one hand, and really radical, structural change on the other, something quite untypical, which may constitute our predicament, but is not normally part of the human condition? 2

2

Cf. Alexander Vucinich: Social Thought in Tsarist Russia. The Quest for a General Science of Society, 1861-1117. The University of Chicago Press, Chicago and London, I W i , X

Foreword When Marxism did eventually come to face this problem, it did not occur primarily through the consideration of pastoral nomadism, but rather that of the famous Asiatic Mode of Production. This form of society, if it exists, contradicts both the sociological theory and the eschatological hopes of Marxism, a number of times over. It is stagnant and self-perpetuating, thus offering no hope to the humanity caught in its toils, unless it be accidental liberation from outside; which, however, must then be contingent on the existence of some other and less stagnant society, and on the conquest of the 'Oriental' society by it. It also offers the spectacle of a self-serving political order built upon violence, and serving the members of the state machine itself and no one else - in other words, a machine of oppression set up not in defence of a pre-existing class system, which has itself been engendered by differential relations to the means of production, but by violence and coercion, and serving a class brought into being by its control of the means of coercion, rather than of production. By allowing coercion to be, in this manner, an independent agent in history, it destroys the optimistic theory that coercion is only a by-product of economic exploitation and can be finally eliminated when such exploitation ends. It thus encourages what Soviet anthropologists have called the 'idealist theory of violence'. This contradiction between the central doctrines of Marxism and the Asiatic Mode of Production has been invoked both against Marxism and against the very idea of the Asiatic Mode of Production. 3

Pastoral nomadism presents a problem for Marxism which is just as fundamental, if less conspicuous. The societies of pastoral nomads are not in themselves offensive and repugnant, notwithstanding the way in which they might appear to hapless populations whom they conquer from time to time. On the contrary, they have often exercised a fascination for outside observers, as objects not merely of investigation, but of admiration. As one Soviet scholar pointed out, here every man was not merely shepherd, but also bard, orator, soldier, historian, senator and minstrel. Nomadic societies know a certain equality (or at any rate a precariousness of fortune precluding stable and internalised inequality), a wide diffusion of civic, political and military participation, an incapsulation of almost the entire culture in each individual, and a certain quite conscious aversion for that division of labour, that specialisation, which Karl Marx also abhorred and wished to see abolished. They often feel and express a certain aversion for the specialist even when they need him, and they relegate him to an inferior 3

There is an excellent discussion of the incompatibility of the notion of the Asiatic Mode of Production with the central doctrines of Marxism in V. N. Nikiforov, Vostok i vsemirnaia istoriia, Nauka, Moscow, 1975. Nikiforov's work constitutes a Soviet answer to the Wittfogel thesis; but Nikiforov agrees with Wittfogel about the incompatibility of the A M P and Marxism. H e disagrees with him only about the question concerning which of the two must give way. For a discussion of Nikilniov's argument, see my 'Soviets against Wittfogel', furthcoming.

xi

Foreword status. If his specialism is religious and requires reverence, this reverence is liable nevertheless to be tinged with ambivalence. Ironically, it is the very attractiveness of nomads which creates a problem. It is not simply that the charm of their society is but one side of the coin, of which the other is their brutality and capacity as raiders and conquerors facts of which the Russians retain a well-maintained historical recollection. The problem does not arise from the fact that they are not altogether attractive. It is rather that as far as theory goes, they have no business to be attractive at all. Their cohesion, egalitarianism, wide social participation, aversion to specialisation, and rudimentary political structures, would be all very well, if only pastoral nomads were still at the stage of primitive communism, or if they had but recently left it. Then one could welcome the appealing traits of nomadic society as survivals confirming that piece of sociological reconstruction. The attractions could be credited to the absence, or at any rate to the but recent emergence, of private property. Unfortunately, the facts of the case firmly exclude such an interpretation. The rough formula which now seems to be accepted (rightly in my view) amongst Soviet ethnographers for the social organization of nomads is this: communal ownership of pasture, family ownership of herds. Moreover, the focus of the major debate - and this was long, persistent and fascinating was land tenure amongst the nomads, not the ownership of animals. The critics of the above formula contended that land was being or had been monopolized by one class within nomadic society. The private, noncommunal ownership of herds was not disputed, even for the past, as far back as the first millennium B.C. If any nomads did indeed pass, as nomads, through the stage of primitive communism, as some Soviet scholars do claim they did, then this stage had to be short and sweet indeed, short enough to leave no traces in archaeological or any other kind of record. This may be the appropriate point to indicate to the reader one of the places where, due to a divergence of conceptual background, Khazanov's work might be misunderstood. Khazanov speaks of inequality amongst pastoral nomads. No doubt there may be a genuine empirical disagreement here with those who find that such pastoralism is conducive to a relative equality; and indeed, the degree of inequality amongst nomads varies according to time and place, and the whole issue is subject to legitimate debate. But part of the disagreement at least is terminological and conceptual rather than substantive. Whether or not concerned with the reconstructed 'stage' of primitive communism, Soviet ethnographic theory is so pervaded by this idea of 'primitive society' - an incomparably more heavily theory-loaded term in Soviet discussions than it is in the West - that it is difficult for a Soviet scholar not to have this notion at the back of his mind, as a kind of back cloth and yardstick, and not to be influenced by it in his choice of language, whether or not he is interested in tluil tillered historic xii

Foreword formation. Where a Western scholar will be struck by the egalitarianism of nomads in comparison with the (to us) repellent extremes of inequality amongst Asiatic sedentary agrarian societies, the Soviet scholar - noting the communalism of nomads, their non-specialisation, rudimentary political centralisation, and collective control of l a n d - can hardly fail to note that, all things considered, nomads are a bit less equal than might be expected, given the lack amongst them of the preconditions of inequality (class monopoly of the means of production) or indeed of its political reflection (developed state formation). Compared with other real, concretely observed large-scale societies, they do seem rather egalitarian, but when silhouetted against the backcloth of primitive communism - which they evoke by their feeble stratification and centralization, and their rudimentary division of labourthey seem less equal than others. Pastoral nomads are a problem for Marxism not merely because of the lack of coherence between their attractive traits and their individualistic economy. As in the case of the Asiatic Mode of Production, it is difficult to explain their political structure in terms of the preconditions or require­ ments of their economic and class organization. Given their private ownership of the crucial animals, they seem to have too little by way of a state, and also too little class stratification. What state formation and social stratification there is, seems to be ephemeral and unstable and elusive. On the other hand, from time to time they seem to acquire far too much of it: nomads have founded astonishing empires. Too much or too little, but never just right: the superstructure does not seem to adjust itself to the requirement of the base, as theory requires. Moreover, if the state oscillates between being under-developed and over-developed, theoretical decency would require that the base keep in step and oscillate similarly (preferably a little ahead in time, as would be appropriate to its causal priority). The available evidence, alas does not confirm any such expectation. Further­ more, the general nature of that superstructure, the social and political institutions of nomads, seems to be often dictated directly by politicalmilitary considerations, by the needs of defence, cohesion and security, rather than being a reflection of the requirements of the social organization of production. Closely linked to all these problems, there is the issue of whether nomadic societies develop. Marxism is profoundly Heraclitean and requires change to be the law of all things: you ought not to be able to immerse yourself in the same society twice, or indeed once. Development towards new and higher forms, through the eventually uncontainable tensions of every classendowed society, is essential both for the sociological mechanics and for the soteriology of Marxism: it guarantees the eventual deliverance of mankind. This is the issue central to Khazanov's treatment of nomads, and the issue xiii

Foreword has a long, complex and fascinating history in Soviet thought, the landmarks of which can only briefly be indicated here. Soviet orthodoxy had at one time endeavoured to incorporate pastoral nomads in the general onward march of humanity, and to some extent still continues to do so. The general device employed for this end was the attribution to such pastoralists of their own special nomadic feudalism. There was, as one nomadic member of a primitive communal society might have ruefully observed to another, a distinctively nomadic way towards feudalism. Was there indeed? One of the paradoxes, from a Marxist viewpoint, of Russian ethnography of nomadic peoples under Czarist rule, was that it was possible to find some who were still semi-patriarchal and not yet properly feudal, and others who had entered market relations and were semi-capitalist and no longer properly feudal, but there was a marked shortage of any nomads who were properly feudal, neither too early nor too late. The two demi-tones were available, but the primary colour in between seemed to be missing from the spectrum. Nonetheless, feudalism was credited to the nomads. The giant amongst the scholars putting forward the thesis of nomadic feudalism was Vladimirtsov (who died in 1931). Academician Vladimirtsov was a scholar with roots in pre-revolutionary Russia. He became interested in the Orient in 1905 as a result of the disastrous Russo-Japanese war, and wanted to study Japanese. The University of St Petersburg, however, was ill-provided with Japanese scholars, although well-equipped with Mongolian ones; so he became a Mongolist instead. Though he noted the need to study Mongol social structure in his diary as early as 1910, he spent most of his life publishing works on Mongolian language and literature, as well as a biography of Ghengiz Khan, a translation of which by Prince Mirsky appeared in London in 1930. But it was only close to the end of his life, in 1930, that he set out to carry out his project, and his crucial book appeared 4

4

B. Ya. Vladimirtsov: Obshchesivennyi stroi mongolov. Mongolskii kochevoi feodalizm. Izdatelstvo Akademii Nauk SSSR, Leningrad, 1934. A l s o translated into French as: Le Regime Social des Mongols: le feudalisme nomade; preface par R e n e Grousset; translation par Michael Carsaw. Paris: Adrien Maisonneuve, 1948. One of the early Soviet ethnographic affirmations of the feudalist thesis is to be found in R. Kabo. Ocherki istorii i ekonomiki Tuvy (Sketches of the History and Economy of Tuva), Gosudarstvennoe Sotsialno-ekonomicheskoe Izdatelstvo, Moscow/Leningrad, 1934. A n important theoretical contribution to the debate also appeared the same year, by S. F. Tolstov, Genezis feodalizma v kochevykh skotovodcheskikh obshchestvakh (The genesis of feudalism in nomadic pastoral societies), which constituted a chapter of Osnovnyeproblemy genezisa i razvilia feodalnovo obshchestva (Basic Problems of the Genesis and D e v e l o p m e n t of Feudal Society), ed. by S. N. Bykovskii and others, O G I Z , 1934. This constituted the proceedings of a plenary session of the state A c a d e m y of the History of Material Culture, which took place from 20 to 22 June 1933. Tolstov was destined to play an important part in Soviet ethnography. His contribution is interesting, amongst other reasons, for his open stress on the political nignilkiincc of the problem. For instance, he affirms in this contribution: 'At present, I lie "Iribnl mirvivnls" XIV

Foreword posthumously. Its subtitle - Mongol nomad feudalism - conveys its thesis. Whether or not all his details have stood the test of further research, and whether or not one agrees with its central idea, it is an impressive piece of scholarship. Moreover, it pre-dates the excesses of Stalinism and contains neither sycophancy nor any evidence of political motivation of the main thesis. If one does not accept his conclusion one may, leaving aside details, invoke the following consideration. Vladimirtsov was primarily an oriental­ ist, and relied on texts above all. Texts tend, however, to stress ideal and legal requirements, rather than concrete social reality. In texts at any rate, Mongols appear subject to complex nuances of unsymmetrical rights and duties between various ranked layers of the population, in a manner which certainly suggests a 'feudal' society. Furthermore, and more significantly, much of the material supporting the feudal interpretation comes from the imperial period of Mongol history, and shows that, at a time when Mongol herdsmen were doubling up as soldiers of the empire, a streamlined military organization was superimposed on the system of clans. The leaders of the decimally organized military units had to be located in prescribed places so that the Emperor could mobilize them, and the ordinary Mongol in turn had to graze his flocks in the area assigned to his superior military officer. This can be made to look like the granting of land in return for military service. In general, the central charge that can be levied against the feudalizing thesis is that it takes its evidence from periods when either the Mongols were conquering, or when they were conquered and incorporated in the Manchu empire. Vladimirtsov's earlier and rather un-doctrinal work does not even cite the classics of Marxism; by contrast, in later years even revisionists quote scripture for their purpose. But his ideas were destined to become crucial on later and much more overtly and deliberately political occasions, notably at the Congress convened in 1954 in Tashkent to consider the nature of social relations amongst the nomadic nations of Central Asia and Kazakhstan. Here the feudalists prevailed, but did not succeed, or were not allowed, to extinguish all opposition. At least one firmly dissenting voice, that of Tolybekov, refused to be silenced. But before we reach this high point in the debate about the feudalizing thesis, one other book, highly relevant to the manner in which Soviet ex-nomads were to be related both to the national question and to the pattern of human history, should be considered. 5

5

emerge as the most dangerous weapon in the hands of the class enemy, aiming to use them at the new stage of the class war in his struggle against the construction of socialism.' Materialy Obedinennoi nauchnoi sessii, posviashchennoi istorii Srednei Azii i Kazakhstana v dooktiabrskii period. Izdatelstvo Akademii Nauk Uzbekskoi SSR, Tashkent, 1955. L. P. Potapov reaffirms the feudal thesis, with special reference to Tuva, in his contribution to a publication which appeared as late as 1975, namely Sotsialnaia istoriia narodov Azii (Social History of the Nations of Asia) cds, A . M. Reshetov and Ch. M. Taksami, Nauka, Moscow, 1975.

XV

Foreword In 1947 the Soviet Academy of Sciences once again published a book of great interest for the history of this debate, namely Vyatkin's Batyr Srym. The hero of the book is the leader of the Kazakh struggle against Russian Czarist imperialism in the late eighteenth century. The book is concerned with the issue of nationality and nationalism, and the analysis of Kazakh nomadic society is only introduced indirectly, in so far as a national struggle, for a Marxist, must also be interpreted in class terms. Vyatkin's formulation of the problem is interesting. Stalin had shown, he notes, that the national problem is in its essence the peasant problem. Now that is all very well, but what do you do when you plainly have a national conflict, but you have shepherds and pastoralists where there should have been peasants? What do you do in a situation in which peasants and serfs are in very short supply? That is the question. Is the national question not merely a peasant question but also a shepherd question? There seems to be no canonical authority for saying so. So what's to be done? Another solution is, however, available. At the Tenth Congress of the CPSU(b) in 1921, Comrade Stalin had provided invaluable help towards the solution of this problem, Vyatkin notes, by characterizing the recent condition of the various pastoral nations of the Soviet Union as patriarchal-feudal. Vyatkin here uses an a fortiori argument: if they still retained patriarchal elements at the time of the October Revolution, they can be assumed to have been endowed with even more of them two centuries earlier! This solution, which consists essentially in seeing nomadic society (or rather, in this case, a single nomadic society at a certain time), as possessing a mixture of patriarchal-communal and of feudal traits, was one destined also to be adopted later in the masterly studies of Kazakh society by Tolybekov. 6

Vyatkin sees and recognizes that under conditions of pastoral nomadism, the maintenance of clan units is essential, and these in turn perpetuate collective ownership of land. So a proper feudalism, in which a class would monopolize land and also deprive ordinary clansmen of their freedom and turn them into serfs, does not develop. Nevertheless, masked by the retention of the ideology of kinship and collectivity, feudal relations do emerge. Vyatkin is sympathetic to his hero and sees him as the champion of the oppressed Kazakhs both against Czarist imperialism and against their own emerging aristocracy, tempted by its own interests into collaboration with the alien conquerors, or at least into dragging its feet in the national struggle. But a proper pastoral equivalent of a Bauernkrieg never developed, simply because neither the leader nor the followers ever freed themselves from the false consciousness of tribalism, which masked an emergent feudal reality. 6

M. P. Vyatkin: Batyr Srym. Izdatelstvo Akademii Nauk SSSR, Moscow, Leningrad, 1947.

xvi

Foreword A theme which is of interest, and which was subsequently to reappear, is the idea that feudalism amongst pastoral nomads is liable to be connected with the imposition of alien domination. It should be added that although Vyatkin wrote and published at the height of Stalinism, and does contain the inevitable canonical vindications of central points, his book is nevertheless very interesting, and is argued at an extremely high level. The next important occasion in the debate occurred in 1954, after Stalin's death, when an inter-disciplinary conference was convened in Tashkent to discuss the pre-revolutionary history of the nations of Central Asia and Kazakhstan, a conference whose proceedings were published the following year. This seems to have been a somewhat openly political occasion, and one which for a time established the feudal and developmental theory of pastoral nomads as the orthodoxy. Its main protagonist was one L.P. Potapov. He castigated the view, prevalent before the Revolution and even during the early years following it, that these populations had lived within a tribal structure in which kin relations prevailed and hence there were no classes or class conflict, as a 'theory . . . used by bourgeois nationalists, rightist opportunists and rigid ideologists and defenders of exploitative classes destined for liquidation amongst these nationalities . . .'. (Italics mine.) Potapov's concern is clearly different from Vyatkin's. Vyatkin en­ deavoured to give a class interpretation to a national conflict which had occurred two centuries earlier. Potapov is, in effect, offering an interpreta­ tion of very recent, post-revolutionary conflicts. If he had allowed that these nomadic nations were made up of kin communities, without anything much in the way of class formation, he would have been faced by a problem. One can imagine a Kazakh herdsman in his pastoral collective, scratching his ear with a bit of stubble as his herd grazes peacefully, and wondering: 'now if we Kazakhs really had no classes to speak of before the Revolution, who exactly was it that we liquidated during the recent decades? A rum business . . .' Potapov's emphatic feudalism provides a clear and unambiguous answer to this conundrum. In his own words, they were 'exploitative classes destined for liquidation'. Nothing could be clearer. At the same time, he rescues nomads for a developmental vision of human history, ensuring that they, and thus humanity at large, were available for eventual salvation, whether or not they happened to be engulfed by an alien imperialism. On Potapov's account, nomads must have passed through, as pastoral nomads, at least three very important and distinct stages - primitive communism, patriarchal society and feudalism. Potapov asserts in so many words that the earlier nomadic pastoral community had shared ownership not merely of 7

8

9

' Materialy xvii

Obedinennoi

nauchnoi se.uii etc.

» Op. cit., p. 17.

9

Op. cit., p. 22.

loreword pasture, but also of herds. His use of the argument from survivals is strange. He admits that the conditions of nomadic pastoralism were unfavourable to collective ownership of herds, that nomadism only began in the first millennium B.C., and that by the middle of the millennium, when historical and archaeological evidence concerning nomads becomes more richly available, communalism is no longer to be found. Yet survivals of it are alleged to persist in more or less contemporary ethnography! This seems to imply that a highly unstable social condition, which could barely have lasted a few centuries before it was displaced because of its internal organisational incoherence, nevertheless leaves social marks which then tenaciously perpetuate themselves for two and a half millennia . . . Whatever the merits of his anthropological ideas, it would be difficult to deny that Potapov was capable of eloquence and vigour in political denunciation. The person whom he denounced with specially forceful irony was S. E. Tolybekov. 'It seems to me, Comrade Tolybekov, that the Khans and Sultans, if only they were here, would receive you with a standing ovation . . .' These feudal lords would welcome Tolybekov's views that they were not true feudal lords after all, because they did not own the land and pasture used by their societies. Starting out from such erroneous theoretical premisses, one can reach erroneous political conclusions, and end up objectively as a defender of large feudal property-owners, Potapov notes pointedly. Tovarishch Tolybekov, Potapov adds darkly, finds himself in just such a danger. The objective dangers of Tolybekov's position (in whichever of the two possible senses one chooses to interpret that phrase) evidently did not intimidate him. Far from it. On the contrary, they seem to have stimulated him into a lifetime of devoted scholarship concerning the social history of the Kazakhs, the theoretical backbone of which is clearly a passionate repudiation of the 'feudal' thesis. His two books on this subject, which admittedly repeat each other in some measure, contain superb and extremely rich ethnography, coherently and interestingly organized around his central ideas. 10

11

10

11

Op. cit., p. 138. S. E . Tolybekov: Obshchestvenno-ekonomicheskii stroi kazakhov v XVII-XIX vekakh. Kazgosizdat, Alma-Ata, 1959, and S. E. Tolybekov. Kochevoe obshchestvo kazakhov v XVII'- nachale XX veka. Politiko-ekonomicheskii analiz. Alma-Ata, 1971. A year before Tolybekov's first book, a volume appeared dealing specifically with the Kazakhs, and still affirming their feudal status, at any rate during the first half of the nineteenth century. It was S. Z. Zimanov's Obshchestvennyi stroi kazakhov pervoi poloviny XIX veka (Izdatelstvo Akademii Nauk Kazakhskoi SSR, Alma-Ata. 1958). But Zimanov was obliged to moderate his thesis: 'The monopolistic right of the feudals to pasture was a factual and not a juridical right.' (op. cit., p. 148.) Here again, feudalism is said to be lurking under a communal, tribal guise, preserving earlier legal forms. Zimanov stresses in particular the role played by the subtle rank differentiation within the two principal, 'real' classes (the rulers and the exploited), in inhibiting the development of an effective class consciousness. The consequence of this was that when genuine social inovcmetttH m Uo ~ and xviii

Foreword One of the most interesting aspects of Tolybekov is his values. He is himself of Kazakh background, and his name is clearly a Russification of Toly Bey. (Amongst the Kazakhs, however, unlike other Turkic groups, bey is an honorific term not implying membership of a hereditary aristocracy.) Unlike his predecessor Vyatkin, Tolybekov repudiates Batyr Srym's rising of 1783, which he insists was motivated by nothing better than the pursuit of loot, slave-raiding, and opposition to the unification of the junior Kazakh zhus (maximal segment) with Russia. Elsewhere, and more than once, Tolybekov insists that the Kazakhs united with Russia voluntarily. (In fact, fear of renewed aggression by the then ascendant Djungarian Mongols may have had some connection with Kazakh eagerness for Czarist protection.) His admission that Batyr Srym and his followers opposed this unification does not amount to a contradiction, in so far as he explicitly says that the beneficiaries of these raids by Batyr Srym were simply the members of the parasitic batyr class, such as Srym himself. (On this point he agrees with S. Z. Zimanov, who otherwise takes a kinder view of Srym's movement. See n. 11.) Srym was, as you might say, a feudal bandit, rather than a social bandit. Tolybekov's warm ex post facto endorsement and ratification of the incorporation of the Kazakhs in the Russian world, does not, so to speak, hang in the air in an opportunistic manner, as a piece of political sycophancy. Tolybekov's retrospective repudiation of the primary resis­ tance of the Kazakhs to Czarist imperialism is, on the contrary, rooted in an important, convincing, and persistently reaffirmed sociological theme: that nomadic society is stagnant. It does not, and cannot, as pastoral nomadic society, develop any further. It constitutes a sociological cul-de-sac, or, to use the expressive Russian word, a tupik. His sociological reasoning as to why this is so, and must be so, is complex, subtle and well documented: it would certainly deserve intensive study and a much more detailed exposition than it is possible to offer here. It also contains some surprising he is willing to class Srym's rebellion as such a movement - in the end they serve not the narod which provides its driving force, but the egoistic interest of this or that sub-stratum of the ruling class, which happened to associate itself with it. (op. cit., pp. 288-9.) H e admits that the feudalism of the Kazakhs was poorly developed, even in the nineteenth century, but connects this with the absence of urban traders, artizansetc. amongst them. (op. cit., p. 290.) This is a theme which also reappears in the work of S. Vainshtein (see n. 18). A similar picture of part-feudal, part-patriarchal and communal society emerges from the major book devoted to the Kirgiz, S. M. Abramzon's Kirgizy i ikh etnogeneticheskie i istoriko-kulturnye sviazi (Nauka, Leningrad, 1971). The author asserts the presence of 'feudal relations' amongst the Kirgiz in the sixteenth and seventeenth centuries, and asserts that these must have emerged no later than the end of the first millennium A . D . But he rapidly goes on to qualify this by saying that the low and backward level of productive force found amongst pastoral nomads, ensured that these features remained intertwined with 'pre-feudal, kin-patriarchal, communal relations' (pp. 155 and 156). Such a formulation is certainly congruent with Tolybekov's views. But Abramzon goes out of his way to polemicize will) L. Krader (Social Organisation of the Mongol-Turkic Pastoral Nomads, Indiana Univ. JAibl., 1963) for exaggerating the importance of corporate kin groups, and missing out the feudal trails (p. 209). xix

Foreword elements, such as the idea that the protracted period of awaiting economic reward amongst pastoralists, which is a consequence of the long gestation period of camels, inhibits the development of productive forces. This is in marked contrast to the Western tendency to see Delayed Return, and the capacity to wait for it, as the prime mark of economic virtue and a crucial factor in innovation - an idea forcefully reintroduced into anthropology by Dr James Woodburn. Tolybekov's repudiation of Kazakh nomadic separatism strictly follows from this theory. On their own they were debarred from any real development. More specifically, and with a great wealth of ethnographic and historical documentation, Tolybekov denies the capacity of pastoral nomads in general, and of Kazakhs in particular, to advance anywhere near a ripe and proper feudalism. No feudalization without sedentarization might well be his motto. The picture of Kazakh traditional society which emerges from his analysis and documentation contains precisely those traits which have led Western scholars to use terms such as 'segmentary' in connection with pastoral nomads: the weak, elusive, ephemeral nature of political centralization, the wide diffusion of power and political participation, the precarious and relatively mild degree of social differentiation, the prominence of collectives practising mutual aid and self-defence. One may speak of a ruling stratum and also of a servile one, but both are very small in comparison with the numerically dominant stratum of ordinary free tribesmen. The number of servile families attached to the household of a ruling khan barely reaches double figures, and suggests the camp of a chieftain, not the court of an oriental monarch. Tolybekov roundly accuses the feudal school of projecting Western or Russian medieval developments onto the Eurasian steppe. Even as conquerors, the Hun or Mongol rulers did not have the capacity for establishing feudalism which was displayed by Teutonic barbarians, who were familiar with the use of serf labour. Tolybekov argues in effect that, compared to central Asia, the West was fortunate in possessing a better class of barbarian, if you know what I mean, who were endowed with a far greater potential for progress. If anyone had developed a mature feudalism in central Asia, it was the sedentary population of Khorezm, but what it had built up was destroyed by the Tatars. Tolybekov invokes S. P. Tolstov who pointed out that it was Khorezm and Kiev which, by their resistance and sacrifice, exhausted the Mongols sufficiently to save Europe from also being overrun by them. It was Khorezmian and Kievan blood which saved Europe, and thus allowed it to develop. 12

13

12

'Hunters and Gatherers today and reconstruction of the past', in E. Gellner (ed.) Soviet and Western Anthropology, Duckworth and Columbia University Press, London and N e w York, 1980. " f'o sledam drevnekhorzmskoi tsivilizatsii, Moscow and Leningrad, I'MH. |>|>,

XX

Foreword Thus Tolybekov unambiguously condemns nomadic pastoral society, when considered from the viewpoint of its contribution to the evolution of human society. It constitutes a barrier to further development. He firmly denies that it can ever reach even mature feudalism - let alone anything beyond that. He firmly rejects the view of his predecessors that underneath the patriarchal and kin terminology of nomads, a feudal society was hiding, wildly signalling to be liquidated. He rather manages to score against his opponents, who claimed that this underlying feudal reality had succeeded in establishing itself surreptitiously, under the guise of patriarchal and clan ideas, and had used this camouflage to befuddle its potential opponents, such as the hapless Batyr Srym. If this doctrine is taken seriously, it seems to imply that a transition from one social form to another can take place quietly, without the aid of violence as the licensed midwife of history; that, in brief, a peaceful transition to feudalism is possible. Tolybekov evidently obtains a good deal of satisfaction from highlighting this implicit heterodoxy of his opponents. Tolybekov himself characterizes nomadic society as patriarchal-feudal, as a transitional stage in which some weak feudal traits appear in a pre-feudal social order. His use of the term 'transitional' (perekhodnye) is strange, in so far as it is absolutely central to his much reiterated position that this transition can never be completed by nomads as nomads. Is a transition that can never be completed a transition? Can a bridge be a cul-de-sac? But if Tolybekov roundly condemns the lack of growth-potential of nomadic society, and welcomes its incorporation in a society which did have the seeds of growth, he is far from hostile to nomad culture, and Kazakh culture in particular. His account of its merits is eloquent and moving. Every illiterate nomadic Kazakh, like all nomads of the world, was in the fifteenth to the eighteenth centuries simultaneously a shepherd and a soldier, an orator and a historian, poet and singer. All national wisdom, assembled by the ages, existed only in oral form. 14

Thus, though Tolybekov does not say so, nomads already in some measure exemplify that multiplicity of roles, that overcoming of the division of labour, that multi-faceted human personality, which Marx in the German Ideology predicted only for the liberated man of the future. Tolybekov in fact warns the Russian reader against the unimaginative philistinism which might lead him to fail to appreciate the beauty of Kazakh lyrico-epic poetry, in which, for instance, the movements of the great Kazakh beauty, Kyz-Zhibek, a kind of Helen of Troy, are compared to those of a three-yearold ram. The failure to appreciate the beauty of such a simile, Tolybekov sternly warns us, only goes with an inability to understand that aesthetics u

S. E. Tolybekov, Obslwlwxtvemw-ekimomicheskii

XXI

stroi etc., p. 426.

Foreword vary with the material conditions of life. Certainly, given a pastoral infra-structure, it is entirely fitting to compare a girl with a three-year-old ram. Even without being a pastoral nomad, I find the idea of a girl moving like a three-year-old ram exciting. Perhaps, one wonders, the Kazakhs actually preferred a three-year-old ram? Whatever the truth about these details of Kazakh culture, and whatever the general terminology he employs to describe it, there can be no doubt concerning Tolybekov's characterization of Kazakh society and of pastoral nomadic society in general. Its culture is widely and evenly diffused and incapsulated in its members, its stratification is ephemeral and weak, its political formations are fragile and elusive, and even if on occasion they grow into something bigger, this leads to no permanent, irreversible, structural changes in society. It is this doctrine, and all it implies in terms of the intellectual tradition within which it was articulated, which is central to Tolybekov's work, and to his life-long struggle with the 'nomadic feudalism' thesis; though he must be at least as much valued for the richness and suggestiveness of his historical and ethnographic material. The two men who continued Tolybekov's argument against the feudalizing thesis, and the recognition of the basically stagnant or oscillating nature of pastoral nomadic society, are G. E. Markov and A. M. Khazanov. Khazanov presents his own case in the present volume better than anyone else could do on his behalf, and there is no need to summarize his position. But the remarkable work of Markov, the present holder of the Chair of Ethnography at the University of Moscow, does deserve mention in this context. Markov and Khazanov (Markov is somewhat older) seem to have reached similar conclusions independently. If Tolybekov's central concern is the Kazakhs (though he firmly generalises his conclusions), the main concern of Markov's book and of Khazanov's present work is comparative (though this is not the case for Khazanov's earlier work on Scythians). The background knowledge they bring to this theme is not identical. Markov refers in his book to many years of repeated seasonal field trips to the nomadic, or rather ex-nomadic, parts of the Soviet Union, and in his book he combines this ethnographic background with a very thorough use of the traditional historical documentation concerning the nomads of Asia (those of Africa being excluded from its purview). Khazanov was initially a historian of the Scythians, and his major previous work deals with them. His concern, in that work, with the cyclical pattern of Scythian history, is then expanded into a general theory of pastoral nomadism, expounded in this volume, and sustained by a remarkable familiarity with modern Western anthropological literature concerning pastoral nomads. 15

15

G. E. Markov: Kochevniki

xxii

Azii.

Izdatelstvo Moskovskogo Universiteta, Moscow, 1976.

Foreword As Markov's excellent study is not available in Western languages, a brief account of some of its points is apposite here. He notes that the origin of pastoral nomadism remains unclear, but that it was preceded by complex non-nomadic agriculture. He locates the emergence of full nomadism at around 1000 B . C . But he explicitly endorses Tolybekov's assertion that in essential social features, the Huns of the third and second centuries B . C . , the Mongols of the eleventh and twelfth centuries, and the Kazakhs of the fifteenth to eighteenth centuries, were similar. However, whilst endorsing Tolybekov's critique of the feudalizing school, and castigating the evidently traumatic 1954 discussions as scholastic, Markov does not endorse the terminology - at least - of Tolybekov's positive solution, i.e. his general characterization of pastoral nomadic society as 'transitional patriarchalfeudal'. He shares Tolybekov's view that evidence for feudal land relations amongst nomads is drawn either from what the Czarist administration had imposed on them, in violation of their own customary law, or (when based for instance on Piano Carpini's reports on the Mongol empire) from temporary devices superimposed by Ghengiz Khan on the Mongol clan organization in the interests of military centralization. He also asserts explicity that amongst nomads, developed forms of class conflict are absent. 'History knows no rising of nomadic tribes, comparable to peasant risings. ' 16

n

Inequality amongst nomads is not very great, and when it does emerge, is a consequence of war and trading rather than of the normal functioning of the economy, within which leaders, members of the privileged stratum, have no interest in depriving their fellow tribesmen of access to the means of production. (One might say that if capitalism requires a reserve army of unemployed, pastoral nomadic chiefs need a reserve army sans phrase; and they can hardly deprive themselves of it, by denying their shepherdreservists the means of subsistence.) Productive relations had the form of economic cooperation or consent, not of feudal dependency. Markov criticises Tolybekov for overconcentrating on the relatively small classes of leaders and servile dependents, to the detriment of the far larger and more typical middle stratum of free tribesmen. It is this stress which enabled 16

17

A recent work dealing specifically with this issue is V. A . Shnirelman, Proiskhozhdenie skotovodstva, Nauka, M o s c o w , 1980. T h e latest contribution to the problem of the socio-political organization of nomads seems to be S. A . Plemneva's Kochevniki srednevekovia (Medieval Nomads), Nauka, Moscow, 1982. This work combines an attempt at formulating a three-stage theory of nomadic society, with a stress o n the impermanence of political formation amongst nomads (a stress specially manifest in a fourfold typology of their political decline and disintegration). This would seem to imply a marked concession to an 'oscillatory', rather than developmental, theory of nomadic political formations. Op. cit., p. 305. The absence of visible or self-conscious stratification in central Asia was a practical and political, as well as a theoretical problem for the Soviets. Cf. Gregory J. Miissell, The Surrogate Proletariat, Princeton University Press, Princeton, N.J., 1974. The book discusses the attempt to see the feminine sex as a whole as the exploited class, for lack of any other plausible and really stilittMiclory candidate for this role. xxiii

Foreword Tolybekov to include the feudal element in his 'patriarchal-feudal' formula. In other ways, Markov is distinguished from Tolybekov by his theoretical caution. Where Tolybekov did have his transitional (with a never-to-becompleted transition) formula for characterizing pastoral nomads generically, Markov declares the problem of the sociological classification of nomads to be unsolved, and thus refrains from attaching any formal label to their social structure. Where Tolybekov asserts with firmness, and perhaps a touch of bitterness, that nomads could not develop further than they did, Markov contents himself with asserting that they did not. Further, he insists that in order to understand their social structure, we must look at the position and role of chiefs amongst them. It could be considerable, in temporary military and imperial situations. But the leaders did not form a closed stratum and did not retain stratum privileges when they lost their leading position. The kinsmen of Ghengiz Khan did have a few privileges, but it did not amount to much. Members of the White Bone clan amongst Khazaks (supposedly kinsmen of Ghengiz Khan) had no great power, even if they attained chiefly status, and were often called chiefs-faineants (mnimymi nachalnikami). They collected no rent and, to maintain their position, were obliged to entertain lavishly, which put a great strain on their households. Markov asserts explicitly that the nomad empires had no economic base. Amongst agrarian populations, government, once it appears, is irreversible, and leads to permanent structural changes; but amongst nomads, centralization, in any case incomplete, is ephemeral, and is followed by a reversal to communal organization. Thus Markov ends with a cyclical or oscillatory account of the socio-political organization of nomads, not unlike that of Ibn Khaldun (who is not invoked by name). But before claiming Markov's conclusions for an Ibn Khaldunian sociology, it is only fair to note certain differences. (These may in fact reflect significantly the differences between the Eurasian steppe on the one hand, and the Arabian peninsula and North Africa on the other.) For Ibn Khaldun, urban life is a permanent necessity, and pastoral tribalism is the only source of state-formation, the state being the gift of the tribe to the ever-present city. Markov's account on the other hand does not make urban life quite so essential as a complementary element for the nomads; and furthermore, sedentary areas seem to be credited with an endogenous and independent capacity to generate political centralization - they do not need to be provided with it by tribal conquest. Finally, it is worth noting the extent to which Khazanov's remarkable earlier work on the Scythians already fomulates the ideas systematized in comparative perspective in this volume. The central idea of continuity in the Eurasian steppe, as opposed to a developmental pattern, is already xxiv

Foreword asserted. In the steppe, Scythians, Sarmatians, Huns, Alans, Khazars, Pechenegs, Turks, Polovtsi, Tatars, the Golden Horde, Kazakhs and others followed each other, but without any basic structural change. He criticises efforts by scholars such as Vainshtein, Artamonov, Griaznov, Chernikov and Smirnov to find a systematic difference between early and late nomads. It is true that the earlier 'houses on wheels' were replaced by yurts that can be dismantled, and that nomads of Iranian speech were replaced by others of Turkish/Mongolian speech; but there has been no fundamental social change. Within Scythian history itself, a certain cyclical pattern emerged. The Scythians established three successive empires, and only the last of these was accompanied by sedentarization and hence by a real state, as opposed to a merely rudimentary tribute-extracting organization of the first two Scythian formations. During this earlier period, the tension between urban and sedentary life and the pastoral nomadic style was already apparent, in the form of hostility between Greek and Scythian. The greater their economic and cultural complementarity, the greater also the political conflict. The Young Scythians, as Khazanov describes them on an implicit analogy with the Young Turks, were notoriously drawn to Greek music and wine (a taste which was of course ruthlessly exploited by Greek exporters), and they delighted in the flute-playing of a Greek captive. But their narodnik King repudiated these Western temptations, and was heard to observe that he preferred the neighing of his horse to that damned Greek flautist. Plus ca change . . . The general conclusions of that study of the Scythians could be summed up as follows: there was urban-tribal economic interdependence, combined with political and cultural tension: political development was cyclical; and social stratification and political centralisation was weak amongst the nomads except at times when they turned themselves into a dominant, conquering stratum of a wider society. Such conclusions can only give pleasure to the present writer, for he is after all a card-carrying Ibn Khaldunian. 18

Fontanili, September 1981 1 8

ERNEST

GELLNER

One of these is accessible in English. Cf. Sevyan Vainshtein: Nomads of South Siberia. Edited and with an Introduction by Caroline Humphrey, Cambridge University Press, Cambridge, 1980. This work of Vainshtein's however is concerned primarily with the ethnography of the Tuvinians, rather than with questions of theory, though the book does contain a brief summary of Soviet debates on these matters. For a recent discussion by numerous Western Scholars of these general problems, see Pastoral Production and Society, collectively edited, Cambridge University Press, Cambridge, 1979. A Western anthropolog­ ist specifically concerned with the central Asian peoples in the Russian orbit is L. Krader. See his Peoples of Central Asia, The Hague, 1963.

xxv

Acknowledgments

The idea for this book belongs not only to its author but also to Ernest Gellner of the London School of Economics. Some years ago after I had published a monograph on the ancient nomads of the Eurasian steppes, principally the Scythians, I considered writing about the historical anthropology of the nomads of the same Eurasian steppes. Professor Gellner prompted me into thinking that it would be more interesting to compare the nomadism of the Eurasian steppes with nomadism in other regions. Of course, this comparison could not be advanced comprehensively by one person. Consequently, I decided to examine one feature which seems to me to be one of the most important for an understanding of nomadism: the interrelations between nomads and sedentary peoples, theoretically referred to in this book as 'the outside world'. While I was writing the book Ernest Gellner generously offered me his invaluable assistance; he read the original synopsis of the book which I wrote in very inadequate English, encouraged me and willingly discussed the work in progress. He even read through the manuscript of the book and made many valuable comments and suggestions which have served to improve the book. I believe that without him this book would never have been written, although, of course, I take full responsibility for its content. When a scholar works in one country and his publisher is in another, when these two countries are separated by different languages, laws and traditions and by other barriers, technical difficulties nearly always arise. Mrs Patricia Williams and Mr Stephen Barr of the Cambridge University Press have admirably coped with these difficulties and done everything in their power to assist me. In many ways their kindness and understanding has made the task of writing this book, while not always an easy one, at least a pleasant one. Thus I would like to express my deepest and most sincere gratitude to them. The transliteration of oriental proper and place names always presents considerable difficulties. Even orientalists use different systems in their xxvii

Acknowledgments transcribing of non-Latin alphabets and literatures, and often not at all consistently. The problem is further complicated by the fact that many historical individuals and peoples appear under different names in the various sources and contemporary works in which they are mentioned. In so far as this book was not written by an orientalist, nor was it written primarily for such specialists, I have tried to keep my transliteration of names as simple as possible and not to use diacritic marks, ligatures and the like. The transliteration of the Cyrillic alphabet into the Latin also involves specific difficulties. The system used in this book was chosen by the translator. My colleagues with whom I worked in the Institute of Ethnography of the Academy of Sciences of the USSR, Dr Krupnik, Professor Pershits, Dr Chlenov and Dr Zhukovskaya, read individual sections of the book's manuscript and made very helpful comments and suggestions. I was able to discuss in great detail with Professor Claessen of Leiden University certain questions connected with general problems about the origin of state and class which are touched on in the book. To all of these colleagues I would like to express my deep gratitude. I would also like to thank my colleagues in other countries who kindly sent me the books, articles and off-prints which I needed when I was writing my book. Principally I must thank the Cambridge University Press and Mrs Williams, also Professor Barth, Professor Bates, Professor Bessac, Dr Bonte, Professor Claessen, Mme David, Professor Dumezil, Professor Ferdinand, Professor Fukui, Professor Gellner, the late Professor Ghirshman, Professor Glatzer, Professor Gr0nhaug, Professor Gunda, Professor Hjort, Professor Jettmar, Dr Komoroczy, Professor von Kussmaul, Professor Malhotra, Dr Masetti, Professor Pastner, Professor Salzman, Dr Sarkozi, Professor Shanklin, Melle Schiltz, Professor Spooner, Professor Stewart, Professor Sweet, Melle Sugita, Dr Szynkiewicz and Professor Van Stone. I owe particular thanks and gratitude to my wife Irina and to my son Yaakov, who did all they possibly could to help me when I was writing this book and who, from time to time, had to resign themselves to the fact that scholarship was depriving them of a husband and a father; to them with love I dedicate this book.

xxviii

Introduction to the second edition

Habent sua fata libelli. The manuscript of the first edition of this book was finished in Moscow at the end of 1979 and was immediately sent to Cambridge. I had many good reasons to hurry. In the worsening political and ideological climate in the USSR, there was no chance to publish the book in the country —it would be considered too revisionist (Khazanov, 1992a). Moreover, it was quite possible that the Soviet authorities might attempt to prevent its publication in the West. Two weeks after I received a message that my manuscript had reached England safely, the Soviet Union invaded Afghanistan. Two weeks later, I ap­ plied to emigrate. Actually, I had decided to try to emigrate several years earlier, but writing this book delayed my application. However, January 1980 was not the best time for such an endeavor. The Soviets never showed good will towards those who preferred the capitalist hell to the communist paradise. In the first half of the 1980s, when detente collapsed completely, they tended to tolerate such people even less than in the 1970s. Just as I had expected, my application was rejected. What I did not anticipate was that it was rejected on the curious ground that my emigration "was not in the interest. . . ." In spite of all my attempts to inquire whose interests the authorities were concerned about-my own, the Soviet state, or maybe the world capitalist system—I never managed to get an answer to this question. Thus, I became but another refusenik with all the consequences. After my application, the Soviet authorities did their best, among other things, to cut all my communications with the West. To a certain extent they failed but that was not their fault. Even the Iron Curtain had some holes. Nevertheless, it took almost five years for this book to be published. During those years the staff at Cambridge University Press was extremely patient, understanding, and sympathetic; they were eager to assist me in all possible ways. I wish to use this opportunity to express my great gratitude to them. Many other people, both my colleagues and those who did not have any connection with anthropology whatsoever, also helped me considerably. Thus, Michael Lavigne, an American who at that lime was living in Moscow, assisted me very much in my correspondence with the West through clandestine channels, alxxix

Introduction

to the second

edition

though this involved a significant risk to his own safety. Neal Naimer, then a young Englishman, specifically came to Moscow to smuggle the proofs that had not reached me by mail. Should I add that all these people occupy a very specific place in my heart? Some beguiling things happened during those years as well. The KGB could not find a better time to search my apartment than at the moment when I was absorbed in work on the book's index, and cards were spread all over my study room. I asked the officers not to disorganize them. At the time the irony escaped me that I was almost repeating the famous plea of Archimedes. To my great sur­ prise, the KGB men behaved themselves better than the Roman centurion from the legend. Not only was I spared, but my cards were not left in great disarray, although the officers spent several hours painstakingly checking whether they contained any anti-Soviet propaganda. Apparently disappointed, they satisfied themselves with confiscating some materials from my archives and all books in Hebrew. This is not a proper place for reminiscences, so, I will cut the story short. The book was published in 1984 and the Soviet authorities, being, to say the least, not particularly helpful in facilitating its publication, hurried to misappropriate my royalties. This was not surprising. The Soviet Communists hated the West­ ern democracies but adored their hard currencies. The same year, though not directly connected to this book's publication, I received the last official warning from the KGB to immediately stop my so-called 'anti-state activities.' However, seven months later, quite unexpectedly, I received permission to leave the coun­ try. I was ordered to depart in a few days time, leaving behind all my archives and possessions. I have many reasons to believe that my colleagues in the West were instrumental in preventing my imminent arrest and in persuading the Soviet authorities that letting me out of the country would be the best way of getting me out of their hair. In particular, I would like to mention my good old friends Henri Claessen and Ernest Gellner. There are many other people that I would like to mention, but I am afraid that I would overlook somebody. Some of them are too modest to even tell me about their activities on my behalf. I cannot ade­ quately express my gratitude to these people. Be that as it may, on July 7th, 1985, my wife, my son and I came to Jerusalem where I immediately accepted an offer of a professorship from the Hebrew University. Meanwhile, my book began a life of its own. It was favorably re­ viewed in several countries (Bates, 1984; Flather, 1984; Leach, 1984; AlAzmeh, 1985; Ecsedy, 1985; Flaes, 1985; Golzio, 1985; Hart, 1985; Humph­ rey, 1985; Ingold, 1985; Ingold, 1985a; Kent, 1985; Oliver, 1985; Tapper, 1985; Wilkinson, 1985; Shimkin, 1986; Tapper, 1986). Some critical remarks notwithstanding, the reviewers usually welcomed most of my main findings and conclusions. The only negative exception was a review by Spooner (1986). The main points of his criticism are as follows. My book does not relate enough to the literature that he, as the specialist in the study of piiHtorcil and nomadic X X X

Introduction

to the second

edition

populations, is familiar with. Sorry, I cannot help him in this respect except to recommend that he read more and keep his mind open. His second point is that the book is more orientalist in nature than anthropological and historical. Again, there is nothing to discuss. This simply means that my esteemed colleague and I conceive anthropology, especially historical anthropology, differently. The third point made by Spooner is that he is dissatisfied with my book in general, and that everything could be done differently and better. In this respect, I com­ pletely agree with him. In my behalf, I can only repeat the old saying: feci quod potui, faciant meliora potentes. For many years I have sincerely hoped that Spooner would do it better; I am still waiting for his own book on pastoral nomadism. In any case, during the last fifteen years or so, multifarious studies of pastoralism came onto the scene with remarkable speed, and valuable new publica­ tions appeared every year. Every time I give a lecture course or a seminar on one or another subject connected with the pastoralists and pastoral nomads, I have the pleasure of updating my syllabi and lists of recommended literature. At the same time, many of my colleagues and students in different countries have indicated that my own book was out of print too long and that its second edition would be desirable. They also pointed out that they would certainly prefer a paperback edition which would make it more accessible. This became especially evident to me after I accepted an offer made by the University of Wisconsin-Madison and, in 1990, joined the faculty there. So, when the Univer­ sity of Wisconsin Press suggested that I publish the second edition in paperback, I temporarily set aside other projects and decided to write this new introduction. In the beginning, I intended to make it a complete review of major trends, prob­ lems, and controversies in the current studies of pastoralism, but, after many pages were written, I had to give up this idea rather reluctantly. I realized that to pursue my original goal further would make this introduction too long. In this case it would be better to write a special survey article, or even another book. Hence, I had to limit myself to an incomplete survey of the most important new publications. Because of the space limitation I am forced to mention as a rule only books and volumes, leaving aside almost all the articles, however important and interesting they are. The subject is becoming so large that it is impossible to give full references to the literature in this introduction. However, I include a brief discussion of a few problems that for various reasons were either neglected in the first edition, or, in my opinion, deserve further elaboration. The book's main argument is as follows: specialization means more depen­ dency. The more specialized mobile pastoralists become, the more dependent they become, in turn, on the outside, non-pastoralist, mainly sedentary world. In this book I concentrate mainly on their economic and sociopolitical depen­ dence, leaving aside their cultural and ideological dependence. However, it is important to keep in mind that puslontlisls, including pastoral nomads, were culX X

HI

Introduction to the second

edition

turally and ideologically dependent upon sedentary societies, just as they depended upon them in economic respects. The economic dependence of nomads on sedentary societies, and their different modes of political adaptation to them, carried corresponding ideological implications. As the nomadic econ­ omy had to be supplemented with agriculture and crafts, so, too, did the nomadic culture need sedentary culture as a source, a component, and a model for com­ parison, imitation, or rejection. I hope to illuminate, at least partly, the latter types of dependency in my forthcoming book: Nomads, Sedentaries and Mis­ sionaries: World Religions in the Eurasian Steppes. In this book the reader will find some classifications and typologies which were accepted by some scholars and criticized by others. I did not expect any­ thing else, and I would like to stress again that definitions, categories, typol­ ogies, and classifications should never become the end of a study, or a purely semantic exercise. They are only its analytical tool, reflecting somewhat the methodological approaches undertaken. Their value depends not so much on their precision, which is almost always disputable, but on how adequately they serve the specific goals of the investigation. The notion of pastoralism advocated in this book is actually based on a con­ tinuum of specific and flexible economic strategies with an almost indefinite range of variations, particularly if different criteria are applied to them. In no way do I wish to make my classification a rigid one, or the only possible one. However, I still hold that such broad typological forms as herdsman husbandry (transhumance), semi-sedentary pastoralism (agro-pastoralism), semi-nomadic pastoralism, and ultimately pure pastoral nomadism as the most extreme form, adequately serve the purpose of this study. They are useful because they imply two opposites: between pastoralism and agriculture and between mobility and sedentism. To make pastoral nomadism, or pastoralism in general, an all-en­ compassing and ill-defined category may be convenient for some archaeologists because of the nature of their sources, but this will hardly advance a better understanding of their prehistoric forms. My classification is primarily an economic-oriented one. First, it is based on the assumption that the pastoralist side of mobile pastoralism determines its characteristic as a specific type of eco­ nomic activity. Second, it focuses on the size and importance of the agricultural component as defining the general subsistence system of a given society. It seems, however, that the correlation between pastoralism and mobility is less clear than I assumed. My classification was challenged by Ingold (1986:165 ff.), who conceives of pastoral nomadism primarily in terms of mobility and resource appropriation and insists that pastoralists have these characteristics in common with hunters-gatherers. (As Ingold, 1986:13, admits, his views on this issue are contradictory. While one should credit his intellectual honesty, how­ ever, nobody but the scholar himself can resolve these problems.) On these grounds Ingold (1980:83, 91) even denies the contrast between the (bod extract­ ing and food-producing economies, as well as the very notion of I ho Neolithic revolution. xxxii

Introduction

to the second

edition

I can only reiterate that sedentism and mobility by themselves do not preclude and define a general economic orientation. Any major form of economic activity has its more mobile and more sedentary forms. Besides, within one and the same mode of subsistence different sectors, groups, families, and individuals may demonstrate different degrees of mobility or sedentism. General Mobility Migrations ( o n e - w a y ticket)

E c o n o m i c Mobility Wandering

1

(hunters and gatherers)

Shifting and slash-and-burn horti/agriculture Pastoral mobility Mobility o f peripatetic ethno-professional groups Industrial mobility ' Maybe the term "wandering" with respect to the hunters-gatherers is not the best one, and I do not insist on it. Still, there is a great difference between their mobility and the mobility of pastoralists.

But for all those to whom the agricultural (Neolithic) revolution is still a valid concept-and the author in spite of his former dissident inclinations certainly belongs to this majority of scholars —one should hardly deny the principal dif­ ference between food-extracting (including hunting) and food-producing (in­ cluding pastoralism) economies, nor the difference between wild and domes­ ticated animals which Ingold tends to underestimate. Although the main characteristic of all varieties of mobile pastoralism, in my opinion, is their specific economic specialization, it would be foolish to deny that this very specialization involves specific lifestyles, world views, cultural values, preferences, and ideals. Pastoralism is not only a way of making a living; it is also a way of living. In this book I do not pay much attention to these, and many other aspects of mobile pastoralism, not because I consider them irrelevant or unimportant, but because every book has its own limits. When one tries to raise all the questions, one faces the risk of becoming superficial, of attempting to say nothing substantial about everything. In spite of the plethora of new literature on various aspects of pastoralism and on individual pastoral groups and peoples, very few books are devoted to the comparative and cross-cultural study of the subject. Besides, most of them deal with contemporary pastoralists, thereby neglecting their historical counterparts. I can only hope that in the future more scholars will heed my long-standing chal­ lenge to combine history and anthropology. Barfield's book (1993) is a good general introduction, or, as the author himself calls it, a textbook, on the subject. The volume published by Galaty and Johnson (1990), which explores various herding systems on five continents, is the broad­ est and most up-to-date survey of contemporary pastoralism. However, the editor's suggested classification of mobile pastoralism in strictly physical geo­ graphical terms seems pioblciiiiiln: (liilniy uml Johnson subdivide mobile pasKxxiii

Introduction

to the second

edition

toralism into pastoralism of the plains (semiarid pastoralism), pastoralism of the desert and tundra (arid pastoralism), and mountain (vertical) pastoralism. In this classification such different types of pastoralism as those practiced in Mali and India, in the Sahara and Botswana, or in the Andes and Morocco, look more similar than they actually are. As a matter of fact, Kyrgyz (Kirghiz) mountain pastoralists in the Pamirs have much more in common with their pastoral Kazakh neighbors, who are typical steppe (plains) nomads, than with pastoralists in Tibet, not to mention in the Caucasus or in the Andes. Two comparative volumes of articles on pastoral nomadism published in Germany (Krusche, 1981; Scholz, 1991) are also worth mentioning. Turning to recent publications on the nomads of the Eurasian steppes, semideserts, and deserts, I should first mention Golden's book (1992). It is an out­ standing achievement and the author is certainly too modest when he calls it just an "introduction." On the other hand, the Cambridge History of Early Inner Asia (Sinor 1990) is not a great success from the conceptual point of view. For exam­ ple, I cannot agree with the editor's strange geographic and cultural definition of Inner Asia. However, the chapters devoted to individual nomadic peoples are, with few exceptions, of high quality. Barfield (1989) has developed further Lattimore's seminal studies of Inner Asian history and contributes to a better under­ standing of the state-formation by the Inner Asian nomads. Other important con­ tributions in this field include works by Beckwith (1987), Jagchid (Jagchid and Van Jay Symons, 1989), and Waldron (1990), Gumilev's (1989) pretentious monograph, published in Russia, on the Eur­ asian steppe nomads, is remarkable only for its unbridled fantasy and ill-hidden anti-Semitism. This is particularly saddening and disappointing since the author was a scion of the two great Russian poets, Nikolai Gumilev and Anna Akhmatova. A more curious general book on pastoral nomadism in the Eurasian steppes (Kradin, 1992) was recently published in Russia. The author is still operating with the Soviet Marxist school of thought and continues to use its terminology, but he honestly points out its numerous deficiencies. In a vain at­ tempt to solve this contradiction he declares that pastoral nomadism is a special socio-economic formation. Ernest Gellner will possibly enjoy this book as more proof that Marxism is still alive in the post-Soviet non-Union (cf. Gellner, 1988); as for me, I do not wish to flog a dead horse. There are numerous historical publications on individual nomadic peoples and polities and their interrelations with their sedentary neighbors and subjects or rulers (Jagchid and Hyer, 1979; Golden, 1980; Dalai, 1983; Vorob'ev, 1983; Egorov, 1985; Morgan, 1986; Allsen, 1987; Halperin, 1987; Jagchid, 1988; Rossabi, 1988; Akhindzanov, 1989; Endicott-West, 1989; Manz, 1989; Kadyrbaev, 1990; Ratchnevsky, 1991; Khodarkovsky, 1992; Kliashtornyi and Sultanov, 1992; Trepavlov, 1993). Ethnographies and anthropological works are less numerous. The most important among them are: Kuzeev, 1978; Vainslilein, 1980; Szynkiewicz, 1981; Masanov, 1984; Zhukovskalu, IWH; Vnlnslitein, xxxiv

Introduction

to the second

edition

1991. With very few exceptions (Humphrey, 1983; Abylgozhin, Kozybaev, and Tatimov, 1989; Bawden, 1989; Zhukovskaia, 1990; Black etal., 1991; and For­ syth, 1992), there are still no objective works on the transformations that the pastoralists underwent during the Communist regimes. Apologetic works, like those published in the past by Soviet and Mongol scholars, or by some Western "fellow-travelers" like Lattimore, can not be trusted. I have good reason to hope that in the future a new generation of scholars in the post-Communist countries will explore this important question much better. The pastoralists and pastoral nomads of the Middle East (Turkey, Iran, and Afghanistan) continue to attract the attention of many scholars, with the pas­ toralists of Iran occupying the major focus (see Oberling, 1974; Ehman, 1975; Stober, 1978; Tapper, 1979 on the Shahsevan; Digard, 1981 on the Bakhtiari; Garthwaite, 1983; Beck, 1986; Black-Michaud, 1986 on the Lurs; Bradburd, 1990 on the Komachi; Beck, 1991 on the Qashghai; and van Bruinessen, 1992, on the Kurds). Several monographs on the pastoralists of Afghanistan describe the situation as it existed before the Soviet invasion of that country (Shahrani, 1979; Barfield, 1981; and Tapper, 1991). The most significant recent studies in history and historical anthropology which are, in one way or another, relevant to the Middle Eastern nomads are: Briant's book (1982) on the pastoralists and pastoral nomads within and on the borders of the empires of the Achamenids and Alexander the Great; a volume on the rise of the Safavids (Savory, 1980); a sketch of Bakhtiari history by Garthwaite (1983); a volume which contains studies on the nomads in Qajar Iran (Bosworth and Hillenbrand, 1983); monographs on the nomads in the Ottoman state in the high Middle Ages (Lindner, 1983), and on the formation and trans­ formation of the gizilbash system (Reid, 1983); a study of the historical develop­ ment of the Qashghai polity (Beck, 1986); and a thesis on the formation of the Indo-Afghan states (Gommans, 1993). Most of the recent studies on the pastoralists and nomads of the Near East are concerned with their changing positions in the modern and contemporary periods (see, for example: Behnke, 1980; Lancaster, 1981; Scholz, 1981; Fabietti, 1984; Anderson, 1986; Janzen, 1986; Lewis, 1987; Hobbs, 1989; Peters, 1990 on the Bedouin and pastoralists in the Arab countries; Marx and Shmueli, 1984; Steward, 1986; Ginat, 1987; Grossman, 1992 on the Bedouin of Israel; Cassanelli, 1982; Samatar, 1982 on the Somali; Casajus, 1987; and Spittler, 1989 on the Tuareg). Of other valuable studies one may single out books on the Rashidi tribal dynasty in central Arabia (Al Rasheed, 1991; see also Vasil'ev, 1982) and on the relation between political forms and Bedouin oral literature (Meeker, 1979). A number of important new publications are devoted to ancient pastoralists and nomads in Arabia and the countries of the Fertile Crescent (Matthews, 1978; Vnrdiman, 1979; Dostal, 1979; Adams, 1981; Eph'al, 1982; Shahid, 1984, 1984a; Anbar, 1985; Parker, 19H6; RiehufU, 1987; Graf, 1989; Finkelstein and X\HV

Introduction to the second

edition

Perevolotsky, 1990; and La Bianca, 1992;) and to the Bedouin in the early Islamic period (Crone, 1980, 1987; Donner, 1981; and Piotrovsky, 1985). The nomads' place in Muslim societies of the Near East is explored in numerous publications by Gellner (see, for example, Gellner, 1981). Their role in the his­ torical process in the region is described in general works on North African state-formation (Berque, 1974; Laroui, 1977; Abun-Nasr, 1987) as well as in more specific publications (see, for example, Berque, 1972; Brett, 1979; Levtzion 1979; Noris, 1986; and Hamani, 1989). On the theoretical level, studies of Near and Middle Eastern nomads are ac­ companied by on-going discussions on the nature of segmentary lineage systems and tribal formations. The debate involves such questions as the degree and forms of hierarchization within segmentary systems and the nature of a tribe: is it a cultural and ideological or political category, or it can be equated with a chiefdom? It seems, however, that the discussion is shifting to the character of interrelations between tribes and states, including the role of tribes in stateformation and vice versa (among numerous publications on all these issues see Eickelman, 1981; Kuper, 1982; Ahmed and Hart, 1984; Crone, 1989; and Khouri and Kostiner, 1990). I should point out that since I am preoccupied, in this book, with the tribalism of pastoral nomads, I do not address the important point that the tribalism in the Near and Middle East was characteristic for some settled agriculturalists as well, though it is true that most of the tribesmen prac­ ticed pastoralism (Tapper, 1990:54; cf. Dresch, 1989). I am more than ready to admit that I consider myself an intruder in the field of African cattle pastoralism. I have never had the chance to do any fieldwork there, or simply to visit that part of the world, and my knowledge of the lit­ erature on the area is limited. In particular, I am not qualified enough to discuss one of the most conspicuous characteristics of pastoralist social organization: the age-set system. Still, I continue to maintain that African pastoralists, like all the others, could not subsist exclusively on pastoral production; that most of them must, in various degrees, be involved in agriculture; and that the emergence of highly specialized pastoralism in East Africa was a rather late development. Most African cattle pastoralists are more egalitarian than their Eurasian steppe or Middle Eastern counterparts, and their lack of a permanent and institu­ tionalized leadership, apparently, may be explained by a combination of several factors. Amongst them I can point to ecological demands; the very limited dispersal of mounted animals, which in other regions served as the backbone of the nomads' military superiority; and last, but not least, the lack of a strong and centralized statehood in the neighboring sedentary areas. Recent works on the African pastoralists embrace different geographical and cultural areas: the Saharo-Sahelian zone (Beckwith, 1983; Dumas-Champion, 1983; Baroin, 1985; Adamu and Kirk-Greene, 1986; Bonfiglioli, 1988), Sudan (Deng, 1978; Kelly, 1985; Robertshaw, 1987; Beck, 1988; Hutchinson, 1988; Johnson, 1989), Ethiopia (Carr, 1977; Almagor, 1978), East Al'rlni (l-'ukui and xxxvi

Introduction

to the second

edition

Turton, 1977; Dahl, 1979; Langley, 1979; Schlee, 1979; Schneider, 1979; Tablino, 1980; Dyson-Hudson and McCabe, 1985; Rigby, 1985; Robinson, 1985; Laube, 1986; Spencer, 1988; Kalinovskaia, 1989; Schlee, 1989). Amongst the newest publications which should be mentioned are the com­ parative volume by Galaty and Bonte (1991) and the highly original, though, to my mind, less convincing monograph by Meeker (1989). In addition to the more or less traditional ethnographic and anthropological studies of African pastoralists, one witnesses an almost explosive growth of publications on their contemporary developmental problems. A very selective and incomplete sample includes: Konczacki, 1978; Raikes, 1981; Sanford, 1983; Evangelou, 1984; Simpson and Evangelou, 1984; Glantz, 1987; Bovin and Manger, 1990; Stone, 1991; Little, 1992, and Baxter, n.d. (see also the numerous publications of various development institutions and bodies). New studies on the Tibetan nomads are few in number (Clarke, 1987; Gold­ stein and Beall, 1989, 1991) for quite understandable reasons. During the last two decades the country remained almost inaccessible to Western anthropolo­ gists. The literature on reindeer pastoralism in Fennoscandia and the Russian North, both in its past and present aspects, is growing (see, for example, Ingold, 1976; Gurvich, 1977; Nickul, 1977; Ingold, 1980; Beach, 1986; Aikio, 1989; Krupnik, 1989; Beach, 1990; Pomishin, 1990; Beach, 1992). For reasons given in this book, the pastoralists of South Asia do not occupy much space my text. Still, some new publications are worth mentioning. Wink's book (1990) contains important information on the nomads' migrations into India in the medieval period (see also Azimdjanova, 1977; Chaudhuri, 1990:138-44, 263-96). The contemporary postures of the pastoralists in India are described in a volume published by Misra and Malhotra (1982) (see also Sontheimer, 1976; Gadgil and Malhotra, 1979; Rao and Casimir, 1982; George, 1985; Salzman, 1986; George, 1990; Gooch, 1992). More information can be found in obscure—by international standards—Indian journals and poorly circulated gov­ ernment reports. Highland or alpine pastoralism (transhumance) in Europe, (including the Mediterranean) as well as in the Caucasus, and the Andes, is practically omitted within this book because it is nothing else, and apparently never was anything else, than a specialized branch of an agricultural economy (for the archaeo­ logical and historical evidence on transhumance and its evolution in Europe see Whittaker, 1988). Braudel was completely right when he noticed that "trans­ humance even in its most disruptive form only concerns a specialized popula­ tion: the shepherds. It implies a division of labor, a settled form of agriculture with crops to maintain, fixed dwellings and villages" (1972:87). In this respect, transhumance in Europe is quite different from the pastoral nomadism and seminomadism practiced in Tibet or the Pamirs. What sets it apart is that when in­ dividuals, communities, or groups of the population in Europe were involved in the pastoralists' occupational specialization, they never constituted a separate XKXVii

Introduction

to the second

edition

society but remained part and parcel of a sedentary society and retained recog­ nized social ties with the agricultural population. They usually shared its culture and its basic values and norms, although sometimes in a slightly altered form. As Barfield (1993:5) aptly remarked: "Heidi is not the story of a Swiss nomad girl even though she herded cows and goats each summer." The llama pas­ toralists are no exception in this respect (of the recent literature see Orlove, 1977; Flores-Ochoa, 1979; Orlove, 1981; Browman, 1987; Brotherston, 1989; Flannery et al., 1989; and Browman, 1990). In the New World, the Araucanian Indians in the pampas apparently repre­ sented a pastoral society—until they were destroyed in general Julio A. Roca's campaigns of 1879-80. New World pastoralism continues to be represented by the Navaho (Navajos), who turned to pastoralism after the Pueblo uprising of 1680 (Underhill 1956:41-3; Vogt, 1961:296; Ellis, 1974:309-24, 481; see also Kluckhohn and Leighton, 1962; Downs, 1964; Terrel, 1970; and Iverson, 1981). I still refuse to regard the Great Plains Indians as pastoralists. Since some scholars have attacked my position on this issue, I feel obliged to explain it at some length. Many scholars would agree that the primary characteristics of the Great Plains Indian cultures that are relevant to the discussion are as follows: 1. Like pastoralist cultures, they developed and functioned in an arid environ­ ment ill-suited for horticulture (Webb, 1931), and, in both cases, the availability of transport animals facilitated their exploitation of vast territories. 2. Unlike the pastoralist cultures of the Old World, they did not originate from indigenous precursors and were not evolving spontaneously, nor did they derive from ordinary diffusion. Instead they grew out of early colonialism. Plains Indian cultures were created by indigenous groups that, pressured by EuroAmericans, had to abandon or transform their previous economies, including horticulture, and readjust to the new physical and political environment (Ewers, 1955:152; Wedel, 1961:284-92)-a readjustment aided by borrowed horses and guns. 3. Except for the dog, the horse was the only domesticated animal kept by the Plains Indians; they had no use for sheep and goats. They kept relatively modest numbers of horses; they had no more than two horses per capita (Barsh, 1990:104). Furthermore, the Plains tribes did not use horse products as food, or at least used them marginally. The backbone of their subsistence economy was bison hunting, but of a type that was very different from pedestrian big game hunting, and that lacked analogues in most other hunter cultures (Murdock, 1968:13-15). One of the few, incomplete, exceptions is the hunting prac­ ticed by some groups of the taiga and tundra zones of Eurasia who kept domes­ ticated reindeer primarily for transport or as a decoy, while their economies remained basically food-extracting ones over the centuries. 4. Although the horse was not directly used for food by the Plains Indians, it made more intense hunting possible. The very presence of the horse introxxxviii

Introduction

to the second

edition

duced some pastoralist elements into the Plains economy, since domesticated animals must be looked after, pastured, provided with fodder, and so on. 5. Though the horse in the Plains Indians cultures acquired a cultural and social value that had analogues in other pastoralist cultures, new social stratifica­ tion and concomitant institutions, such as militarism, may not be ascribable to its introduction alone (cf. Ewers, 1955). The general historical circumstances in the area, including various influences from the Euro-Americans, must also be taken into account. Whether one calls the Plains Indians mounted hunters or nomadic pastoralists largely depends on one's concept of pastoralism. If one agrees with Oliver (1962a:35) that pastoralism is a sociocultural system, a life-way, then one will justly pay particular attention to its similarity to the hunting on horseback, or mounted hunting, of the Great Plains Indians. Likewise, if one follows Ingold (1986:167), who tends to stress mobility as a key element of pastoral nomadism, the difference between pastoralism and mounted hunting may become blurred. However, to me pastoralism is a distinct type (or rather types) of food-producing subsistence economy, and in this respect it is quite different from mounted hunt­ ing. Plains Indians were not typical hunters; nevertheless, they remained hunt­ ers. Therefore, the general orientation of their food-extracting economy was dif­ ferent from that of the pastoralist food-producers. Correspondingly, there are sharp differences in the nature of the periodic movements of mounted hunters and pastoral nomads, and in the determinants of their annual economic cycle (Oliver, 1962). Basically, bison herds move unpredictably (Barsch, 1990:14, 15), although sometimes they were driven by the Indians for slaughter. More­ over, the reasons underlying similar cultural elements in the societies of mounted hunters and pastoral nomads sometimes differ. For example, whereas the Plains Indians deliberately burned the grass to facilitate the bison harvest, the nomads of the Eurasian steppes used this practice not so much for hunting as to increase the pasture for their domestic stock. For these reasons, it seems to me that suggesting that the Great Plains Indian cultures practiced a form of pastoralism overestimates the pastoral side of their economy. Ingold (1986:167) claims he senses that the reluctance to accept the nomadism of hunters and gatherers is an anthropological reflection of the very general con­ tempt that pastoralists have for their hunter-gatherer neighbors, even with cor­ responding moral connotations. In scholarly works it is always better to rely on facts than on senses which often may be deceptive, and I am afraid that Ingold fights the windmills. Neither I nor, to the best of my knowledge, any contem­ porary anthropologist has any bias against the hunters; moreover, Ingold does not substantiate his claim that pastoralists held hunter-gatherers in contempt, or, at any rate, that their contempt was greater than that of sedentary agricul­ turalists. Robertshaw (1989:211) has correctly pointed out that we should be aware of the possibility that the p;isioralisls' contempt for hunting may some­ times have rubbed off onto the cilio*aphcr. In any case, the situation in East

Introduction

to the second

edition

Africa is far from being a universal one; and even there the contempt for hunting is connected with social factors much more than with purely occupational dif­ ferences. In the Eurasian steppes, as well as in the Near and the Middle East, hunting has always been held in high esteem by the pastoralists. Recent studies, particularly in the Levant and adjacent areas, but also in some other regions, have advanced our knowledge of incipient phases of animal domestication and husbandry, prehistoric forms of extensive pastoralism, and the spread of pastoralism on the oikumene (see, for example, Shnirel'man, 1980; Clutton-Brock, 1981, 1981a; Gauthier-Pilters, 1981; Rowton, 1981; CluttonBrock and Grigson, 1983, 1984; Gilbert, 1983; Grigson and Clutton-Brock, 1984; Mason, 1984; Rindos, 1984; Clutton-Brock, 1987; Rosen, 1988; Russel, 1988; Browman, 1989; Clutton-Brock, 1989a; Helmer, 1989; Shnirel'man, 1989; Zagarell, 1989; Gautier, 1990; Hemmer, 1990; Pomishin, 1990; Zarins, 1990; Cribb, 1991; Sadr, 1991; Bar-Yosef and Khazanov, 1992). Apparently, from the biological and cultural perspective, the domestication process involved the following stages: choosing (a suitable species), capture, isolation, taming, behavioral control, controlled breeding (morphological changes), diffusion (tak­ ing animals into new habitats), and adaptation to a new environment (including hybridization). Recent studies in the prehistory of the Levant confirm my conclusion that animal husbandry there did not develop out of hunting and that the domestication process was initiated and conducted not by hunters but mainly by incipient culti­ vators (Helmer, 1989). The evidence for plant cultivation prior to animal domes­ tication is growing. It is true that prehistoric hunters-gatherers, just like their historical counterparts, were capable of taming many animals. From ethno­ graphic data it is known that they liked to keep various mammals, birds, even reptiles as pets. Tamed animals were usually kept for pleasure and sometimes for utilitarian reasons. However, in many important respect, taming is quite dif­ ferent from domestication, which implies a regular reproduction of domestic animals and, eventually, changes in their genotype through natural and, par­ ticularly, artificial selection. Domestication demanded at least three precon­ ditions: (1) a good knowledge of the behavior of the animals to be domesticated; (2) a relatively sedentary way of life; and (3) the disposal of surplus agricultural or vegetal products that may be used for fodder. The hunters, particularly specialized hunters, certainly met the first precondition, but, as a rule, hardly the second and the third ones. Still, in all fairness, I cannot completely exclude the much less plausible possibility that in some cases domestication took place in communities which practiced intensive wild vegetal resource exploitation and storage along with specialized hunting and trapping. For example, some schol­ ars insist that the cultivation of plants did not precede camelid domestication in South America (Browman, 1989; McGreevy, 1989). On the other hand, the ethnographic and historical evidence Hiipporls (he more xl

Introduction

to the second

edition

plausible premise that hunters could shift to pastoralism after obtaining already domesticated animals. It is still less clear whether it was enough to borrow only the "idea of domestication," although some scholars consider this a plausible hypothesis. Thus, it is widely believed that pastoralism was the earliest form of the food-producing economy in the Sahara and the domestication of cattle there was a local phenomenon developed by Epipalaeolithic groups who did not prac­ tice agriculture (see, for example, Muzzolini, 1983; Banks, 1984; Holl, 1989; Clutton-Brock, 1989). Still, one should hardly overestimate the extent to which the borrowing of domesticated animals and particularly the idea of domestication by the hunters led to the spread of pastoralism. Remarkably, the ethnographic materials prove that hunters most willingly borrowed those animals, like dogs, horses, reindeer, and even camels, that could be used as decoys for hunting or for transportation, and thus made more efficient their traditional mode of sub­ sistence (Shnirel'man, 1980:216-7). An even less plausible possibility is that pure pastoral nomadism already existed in prehistoric times. Despite the growing sophistication of archaeological methods, in most cases we still cannot convincingly discriminate between seasonal camps which were left by pastoral nomads or semi-nomads, and those which belonged to groups who practiced a settlement-based transhumance (herdsman husbandry) in which only a part of the population carried out more or less specialized pastoral occupation. Moreover, there is always the risk that archaeological data on a specialized segment of society may be mistaken for the higher taxonomic unit, particularly if some cultural specifics are involved. Ethnographic materials may serve as parallels and as comparative data in our models of the prehistoric past, but they should not be taken as direct analogies for archaeological reconstructions. Prehistorians often prefer to use contem­ porary data, such as recent ethnographic analogies, instead of historical data. The application of ethnographic analogies of pastoral systems requires particular care in the Near and Middle East, and I suspect, in sub-Saharan Africa as well, given the variety of systems found ethnographically even within the same en­ vironmental zones. For example, Sherratt (1981, 1983) provides serious argu­ ments in support of his hypothesis that secondary products of sheep, goats, and cattle, including milk, hair, wool, traction, and pack transport, began to be utilized intensively only in the 5th and the 4th millennia B.C. It is quite possible that only after that event did pastoralism begin to acquire a resemblance, how­ ever incomplete and sketchy, to the forms known in history or used in our ethno­ graphic analogies. All these differences may be irrelevant if one agrees with the point made by the archaeologist Cribb (1991:16) that "the search for a fully nomadic society should be abandoned in favour of an approach which recognizes nomadic tendencies manifested in varying degrees in a wide range of societies and com­ munities." We should not forget, however, that an elimination of the problem does not mean its solution, Kli

Introduction

to the second

edition

I still think that pure pastoral nomadism with the large-scale use of mounted horses and camels must have had very specific preconditions and causes and taken a long time to incubate and to evolve. It was a late development in the Eurasian steppes and an even later one in the Near East (cf. Zarins, 1989). Sig­ nificantly, neither Sumerian nor Akkadian languages had a genuine equivalent for the term "nomad," because at that time pastoralism was still much less specialized, with many tribes or groups having both nomadic and sedentary sec­ tions (Rowton, 1981:28). Although one cannot exclude that the horse and, possibly, even the camel began to be used for riding earlier than it was thought in the past (Anthony and Brown, 1991), large-scale riding was a rather late phe­ nomenon. Still, there is no conclusive evidence to insist that regular riding pre­ dated the wheel transport. New data on the origins and spread of pastoralism in Africa are summarized by Smith, 1992 (see also Webster, 1979; Spear, 1981; Ehret and Posnansky, 1982; Clark and Brandt, 1984; Robertshaw, 1989; Maggs and Whitelaw, 1991; Sadr, 1991). Still, many important questions remain debatable. One of them is whether domesticated cattle, like ovicaprids, were introduced to North Africa and the Sahara from the Near East, or whether they were domesticated in situ, after the concept of domestication became familiar in the region. Another impor­ tant question is whether a direct transition from hunting to herding really took place in the Sahara, and even more so in some other regions of Africa. The third as yet unsolved question is how widely domesticated animals and pastoral prac­ tices were diffused throughout Africa prior to the spread of agriculture. Argumentum ex silentio is certainly not the best proof. Thus, Robertshaw and Collet (1983) hypothesize that the Neolithic population in East Africa actually practiced a mixed economy and that the absence of seeds from domesticated plants on their sites may be due to a number of factors. The Early Iron Age farmers' migration into the Rift Valley (A.D. 200-300) gave the pastoralists their first chance to ob­ tain grain without having to grow it themselves and, thus, around 500 A.D. they began to specialize on herd management (cf., however, Marshall, 1990). The origins of genuine pastoralism in South Africa is apparently an even later devel­ opment (Denbow, 1986). One pressing problem, which for several reasons was not given its proper due in my first edition, is the fragile situation of contemporary pastoral nomads and mobile pastoralists in general. Many valuable contributions to this issue have been published during the last twenty years or so (see, for example, Salzman, 1980; Galaty, Aronson, Salzman, and Chouinard, 1981; Raikes, 1981; Salz­ man, 1982; Sanford, 1983; Simpson and Evangelou, 1984; Bovin and Manger, 1990; Galaty and Johnson, 1990; Galaty and Bonte, 1991; see also numerous publications of the Institute of Development Studies, University of Sussex, the Scandinavian Institute of African Studies, and many others). This makes my at­ tempt to provide a brief and very impressionistic survey of tllO problem much easier. xlii

Introduction

to the second

edition

To summarize, one may say that the current, more often than not miserable, positions of the pastoralists are directly connected to their changing relations with the outside world. In this respect, their past was unique, their present is precarious, and their future is dubious. Some would argue, perhaps hopefully, that they do not have any future. In the past the mobile pastoralists in many regions possessed some military and social advantages in their relations with their sedentary counterparts. These commanding advantages often allowed them to overcome the deficiencies of their specialized but subsistence-oriented economy by non-economic measures, by transferring these interrelations from a purely economic plane to a political one. In other words, their military superi­ ority and their underdeveloped social division of labor turned out to be a leverage of political strength. This was particularly true with respect to the Great Nomads of the Eurasian steppes and the Near and Middle East, where every male nomad was a warrior and most of them were mounted warriors. Pastoral societies began to decay at the onset of modern times, when sedentarists made improvements in transportation and warfare. Caravels, and then steamboats, proved to be more efficient than caravans; and regular armies of sedentary states increased their use of gunpower that was accurate over long distances, and became stronger than irregular cavalries (cf. Headrick 1981). The continental empires of Russia, China, and Ottoman Turkey were the first to en­ croach upon the territories of mobile pastoralists. Others followed in their steps. These changes became evident in some parts of the world earlier than in others, but eventually the process took the same path all over the world. First, the pastoralists lost their military superiority, then their political independence; afterwards they had to adjust to forces outside their control, including the economies of the modern or modernizing sedentary world. The increasing responsiveness to, and dependence on, colonial powers, national governments, and the outside world in general, all of which remain alien powers to most pastoralists, had several detrimental effects. It decreased the size of the territory occupied by them, undermined their subsistence-oriented economy, and eroded the stability of their society. In many countries, the intrusion of farmers into pastoral areas was due to the expansion of the agricultural population and to deliberate anti-pastoralist policies of colonial powers and national governments. For instance, in Central Asia, long before the Bolshevik revolution, the Russian government took away the Kazakhs' summer pastures, and sometimes even their winter quarters, and replaced them first with Cossack and then with Russian peasant settlers (Demko, 1969). About one and a half million new colonists from European Russia came to Kazakhstan at the end of the 19th century and in the beginning of the 20th century (Dakhshleger, 1965:51) Kazakh pastoral nomads were gradually ousted to the arid areas of Central and Southern Kazakhstan. Later, the so-called "virgin lands campaign" of the 1950s, aimed at sowing wheat on huge tracts of land in the steppes of northern Kazakhstan, brought another 1.5 to 2 million new settlers from the European pari of the USNH to (he i.ouuliy. During this campaign the xliii

Introduction

to the second

edition

livestock-raising farms were closed, but most of its Kazakh employees were prevented from becoming involved in grain production. This also happened outside of Central Asia. In southern Mongolia, the Chinese peasant colonization pushed the nomads to marginal lands and even­ tually made them an ethnic minority in what had been their own country (Lattimore, 1940). In South Africa, Swaziland, Lesotho, Botswana, and Zimbabwe, colonial rule not only limited the amount of land available to indigenous pastoralists but it also put a brake on their movements (Silitshena, 1990; Smith, 1992:218 ff.). In Kenya, the Maasai and other pastoralists lost a considerable part of their pastures during and after the colonial period. The loss was caused, first, by white settlers, and then by the gradual expansion of indigenous agriculturalists into Maasailand, and by the creation of wild game reserves (Hjort af Ornas, 1990:97 ff.; cf. Bernus, 1990:167 ff.; Bovin, 1990 on the Sahelian zone; or Loiske, 1990: 82-83 on Tanzania; for a general summary of the situation in Africa see Galaty and Bonte, 1991a). A similar policy was pursued by many governments of the Near and Middle Eastern countries. For example, in Iran the Land Reform Law of 1962-63 pro­ claimed the non-cultivated land, including pastures, to be registered in the name of the state (Beck, 1986:157, 251 ff.; Black-Michaud, 1986:209-10; Bradburd, 1990:21 ff.). In Syria, in 1958, the government declared that the whole un­ cultivated steppe and desert areas were public property and brought agriculture into areas occupied by the pastoralists (Bahhady, 1981:260-1; Lewis, 1987:186; Shoup, 1990:198 ff.). On the other side of the world, in the Scandinavian Arctic and Siberia, many reindeer pastures were lost to northern agriculture and particularly to the hydro­ electric development and extractive industries (Paine, 1982; Savoskul and Karlov, 1988:166-8; Morris, 1990:198-202; Beach, 1990:261-2; Beach, An­ derson, and Aikio, 1992:60 ff.; Forsyth, 1992:402-3). The situation with rein­ deer pastoralism in Russia is much more tragic than it is sometimes assumed, despite the claims that heavy industrialization of the Russian North did not en­ danger the existence of herding practices and that the state-owned reindeer farms are praiseworthy (Beach, 1990:293). Actually, the opposite is true, and at pre­ sent not only reindeer-herding but even those ethnic groups who practice it are in danger. The natives in the Russian North became unprivileged and powerless bystanders. The life expectancy of the indigenous peoples of Siberia at present does not exceed 43 years. This is 18-20 years lower than Russia's average and 29 years lower than the life expectancy of the Eskimo in Canada and Alaska. In the Soviet Union the industrial development of the Northern habitat and the Soviet government's policy towards the indigenous population of the North led to ecological destruction and to the disruption of traditional means of sub­ sistence. Hunting and fishing are in decline, and the reindeer population (1,800,000 head) is at a record low for the 20th century (Pika and Prokhorov, 1988:78, 80). xliv

Introduction

to the second

edition

As a consequence of industrial intrusion, the pasturelands in the North de­ creased 20 million hectares; this left 100,000 reindeer without grazing territory. In the Magadan oblast' (province) the number of reindeer decreased from 1,000,000 in the early seventies to 600,000 in the late eighties. In the KhantyMansi autonomous district the number of reindeer decreased from 71,100 in 1930 to 52,200 in 1982, and since then their numbers have continued to decline (Sokolova 1990:158). A significant part of the pastures in the Taimyr peninsula has been literally burned out by the poisonous gas and acid rain—byproducts of the Noril'sk enterprise. In addition, the gas-line Mossoiakha-NoriFsk broke the migratory routes of the wild deer, which then had to encroach upon the pastures utilized by the domesticated reindeer. Metal and kapron fences harshly built in the tundra in the 1980s resulted in overgrazing and in the further deterioration of the pastures (Tarasov, 1993:6). Even though gas production in the Yamal peninsula has not yet started, the pastures there have already been mutilated by the wheels and caterpiller tracks of heavy wheels. Each oil test boring leaves 4 to 9 hectares of spoiled land. By the year 2000, the pasturelands in the Yamal may shrink by half, at a minimum (Leibzon, 1992:4). The Soviet regime tried in the North to pursue its policy of enforced sedentarization and to model indigenous subsistence-oriented systems after the stateowned agricultural and stock farms. The government ordered the natives to resettle in big settlements where they constitute a despised and discriminated minority among the Russian population. A significant part of them became un­ employed, at best consigned to unskilled and heavy manual labor. By the end of the 1950s, more than 70 percent of the indigenous population in the Russian North were still involved in reindeer pastoralism, hunting, and fishing; by the end of the eighties, only 43 percent were involved in these activities. Less than 10 percent of the natives continue to live the nomadic life; most of them are shepherds who for many months have to live alone with the herds, far away from their families (Pika and Prokhorov, 1988:77). In addition, the state-imposed practice of sending the native's children to the distant boarding schools, against their parents' will, has exposed them to acculturation and alienated them from their mother-tongues, ethnic cultures, and traditional occupations. The catastrophic decrease in the reindeer stock, lumpenization, and mass un­ employment amongst the native peoples of the Russian North destroyed their traditional way of life, their value system, and their dietary system, which dis­ ruptions have led to severe health problems, poverty, widespread alcoholism, and high rates of suicide. Another result has been the inability of many to create a family: many are doomed to remain single. Several years ago the Russian government admitted, for the first time, that the Northern pastoralists faced a catastrophic situation. However, it seems that it is still paying only lipservice to this problem. In spite of numerous conferences and similar measures which smack of a propaganda campaign, a serious desire to address the real issues does not exist. Some experts claim that the money spent on this propaganda would be enough to provide, the Northern aborigines with decent dwellings and utenK I V

Introduction

to the second

edition

sils. Meanwhile, the situation continues to deteriorate. In summer 1993, the pastoralists in the Taimyr for a month and a half could not get even a piece of bread for their children (Tarasov, 1993:6). Pastoralism cannot survive nowadays in its traditional forms. Therefore, some kind of modernization of pastoralism is inevitable. The term "modernization" is currently out of fashion and is substituted by vague euphemisms like "develop­ ment" or "change" only because in the 1950s and 1960s some advocates of mod­ ernization adhered to the "convergence" theory and thought, or hoped, that all modern systems would intimately develop similar major institutional features. To their mind, this should inevitably lead to liberal democracy. Many theorists today claim that this insistence on following the Western pattern of socio­ political development was a fallacy. However, I still think that there is nothing wrong with the term itself, if devoid of this ideological burden. At the same time, we have learned from the many failures and shortcomings of moderniza­ tion that the process involves a deep transformation of the existing social, economic, and cultural institutions which may facilitate or hinder the modern­ ization process. Recent failures in attempts at modernization in many Third World countries have proved that the process entails many difficulties, particu­ larly if specific ways and models of modernization are imposed from the outside. Merely destroying and supplanting traditional forms of social and economic organization does not necessarily assure the development of a new, viable modern order. On the contrary, this often leads to disorganization and chaos (Eisenstadt, 1973, 1983). Modernization remains an alien concept for most pastoral nomads and mobile pastoralists in the Near and Middle East, Central and Inner Asia, Africa, and several other areas, in spite of the fact that the direct and intentional forces of modernization in the form of government planning and programs are often specifically aimed at the pastoralists. It is clear that when modernization is primarily conceived in terms of constant technological innovations and eco­ nomic growth, it is not applicable to mobile pastoralists — until they become a specialized but integrated part of a modern society, with a stake in the wider distribution of its benefits. This problem is at present not just a scientific one; like all problems connected with development, it has become a hot political, economic, ecological, and humanitarian issue as well. Not only anthropologists, but also sociologists, economists, planners, experts in development, politicians, international and governmental offices, and many others are involved in its solu­ tion, often with disastrous repercussions for the pastoralists themselves (Khazanov, 1990). The economic aspect of the problem is as follows. Ecological, economic, and social foundations of traditional mobile pastoralism make it incapable of longterm stable economic growth, based on increases in productivity. In the past, pastoralist economies were never deliberately profit-oriented or consistently aimed at meeting market demands. Instead, they were aimed at subsistence, alx Iv i

Introduction

to the second

edition

though they were hardly ever self-sufficient. However, at present, the pastor­ alists are increasingly being drawn into regional or international systems based on a monetary economy and increasing surplus production. Subsistence-oriented economies, even specialized ones, are easily overstressed when they become dependent on market transactions and are forced to produce in accordance with surplus market demands. Stress occurs because the traditional technologies of a subsistence based economy cannot keep up with the demands of a surplus market. However, the stress is even greater if the sub­ sistence technologies undergo some kind of modernization but continue to operate within the framework of traditional social organization and land tenure. It is difficult for traditional economies not only to compete with, but even to be adapt to, the modern economies. Therefore, it is difficult, in principle, to main­ tain traditional mobile pastoralism within the modern economic climate. Those who are involved in this effort must either modernize it, make it even more specialized along the lines of commercial production, or, on the contrary, make it more diversified and supplement it with some modern activities. Otherwise, they face the risk of being further marginalized and encapsulated (like the Bed­ ouin in Egypt; see Hobbs, 1989) or becoming zoo groups, attractions for roman­ tics and tourists for whom the allegedly eternal and unchanging pastoralists represent "the other," the inventory of a living museum (see, for example, Lavie, 1990 on the Mzeina Bedouin of the South Sinai during the Israeli occupation). Under these circumstances, social consequences may be destructive in any case. Meanwhile, mobile pastoralism in its traditional and particularly pure no­ madic forms, as it existed for several thousand years and the remnants of which some of us can witness in our fieldwork, is, if not already completely dead, then dying. It is dying because it has proved to be incompatible with modern, in­ dustrial, society. And just as traditional societies are more and more transform­ ing into modern or modernizing ones, or are moving further into their orbit, traditional pastoral societies are increasingly transforming into societies of dif­ ferent types. I do not see any long-standing tendencies that can seriously slow down this process, or, even more so, turn it back, although a radical transforma­ tion from one way of life to another is always very painful and often not very successful from both the economic and social points of view. How, then, should we approach the traditional mobile pastoralists in our Brave New World and what model of modernization is the most suitable for them? To my mind, the very formulation of the problem in these terms is wrong. Before any developmental policy from an outside agency is implemented, there should at least be dialogue and negotiation with the population who is being assisted. One should ask not what to do with the pastoralists but what the pastoralists themselves are willing and have to do to adjust to the necessities of modernity; only in this connection may one ask how the pastoralists can be helped in this very painful process. However, governments and international agencies often operate in a different way; (hey prefer to talk and to decide among xlvii

Introduction

to the second

edition

themselves, but not to delegate decisions to the immediate participants of the developmental process. It is true that sometimes it is difficult to organize local participation. Still, the underlying reasons of the decision-makers, even when they are not connected with specific political interests, are invariably the same and are based on two main assumptions. First, in our modern times (should I already call them post-modern?) the mobile pastoralists simply can no longer maintain their traditional way of life and their traditional subsistence-oriented economy, and will have to adjust. Second, the pastoralists themselves are in­ capable of developing any viable adjustment strategies; therefore, they should be paternalized and guided. The first assumption seems to me correct, the sec­ ond wrong. Up to now two major solutions have been suggested, experimented with, and in many cases considered disappointing. The first one, advocated mainly by ex­ perts from Western countries, has as its declared aim a transformation of tradi­ tional pastoralists into commercial livestock producers or even capitalist-type ranch-owners. With few and incomplete exceptions, different ranching schemes and projects turned out to be inadequate in many desert and semi-desert areas, for obvious environmental reasons. Precipitation there is often unpredictable, most of the pastures are seasonal, their productivity is low and considerably variable from year to year. The proper utilization of their forage resources by the pastoralists requires that they should have the freedom to move with their stock over a large territory. No wonder that a policy favoring the privatization of grazing lands or their restrictive allocation to specific groups, like attempts to introduce a ranch-like system undertaken in several African countries, at best achieved only a very limited success (Raikes, 1981; Sandford, 1983; Simpson and Evangelou, 1984). Thus, in Kenya, the practice of granting individual title to huge tracts of land formerly held by the Maasai, in order to develop them as "progressive" beef ranches embracing the principles of capitalist national development, was halted once it was realized that there simply was not enough land to go around. This practice was actually reducing the viability of Maasailand as a whole in terms of its usefulness in supporting the pastoral population. Thus began the present day policy of registering group ranches in which several families form a cor­ porate body and as a business venture obtain title to land. However, this policy conflicts with the traditional way of seasonal utilization of pastures. At the same time it put many pastoralists at a disadvantage and increased social tension; they witnessed common pastures being appropriated and diminished by their more successful tribesmen, or even other people. The process of turning "pasture" into "land" has been not only the undoing of the traditional way of life; it is also a formidable threat to the Maasai collective existence (Galaty, 1980:161 ff.; Evangelou, 1984; Halderman, 1987; Bennett, 1988:49 ff.; Graham, 1988; Bekure and Pasha, 1990; Hjort af Ornas, 1990:98 ff.; Galaty, 1997:26 ff.; Lit­ tle, 1992). xlviii

Introduction

to the second

edition

Another problem is the fact that administrators and planners who tried to turn pastoralists into commercial livestock producers have not realized that produc­ tion is a socially and culturally constructed activity as well as an economic one. Thus, they tend to ignore the peculiarities of the social organization and land tenure of traditional pastoralists, particularly their view that land has no market value and does not constitute private property. Finally, many livestock projects of the 1960s and 1970s were influenced by Western economic theory and assumed that appropriate economic-technological inputs would automatically yield desirable economic outputs. However, many pastoral groups, particularly in Africa, are reluctant to produce for the market because stock for them are not only a means for subsistence, but also a form of wealth, a social capital, and a source of prestige and esteem connected with specific cultural values. Besides, these projects implied that livestock prices would provide the incentive for pastoral production on a commercial basis. However, national governments which set producer and consumer prices on animals and animal products frequently prefer to keep these price levels arti­ ficially low to please their growing urban populations (Raikes, 1981:189; Ben­ nett, 1988:52-5). The open range system that developed in the western half of North America in the last quarter of the 19th and in the first half of the 20th centuries, and in some other countries like Argentina or Australia, was from the outset different in many respects from the traditional pastoral economies. Above all it was quite different in its land tenure and the fact that it was operating within a framework of a capitalist profit-oriented economy (Atherton, 1961; Frantz and Choate, 1966; Gilles, 1987). "Livestock was produced for sale from the very beginning; there was no introductory period of subsistence production, as there was in the Middle East, Central Asia, and Africa . . . North American cattlemen, despite their romantic traditions, were businessmen from the beginning" (Bennett, 1985: 90). All this allowed the open range system to quickly turn into an intensive pro­ duction system with irrigated pastures, machinery, tame forage, breeding, fences, and so on. These categorical differences make it impossible to turn traditional pastoralists into capitalist ranch-owners without drastically changing their social organiza­ tion, destroying their communal forms of land tenure, depriving a large number of pastoralists of free access to pastures, and generally increasing their economic and social differentiation. In many such cases the social cost may be very great. The negative effects of land privatization are most evident when participation in this process and its rewards are not balanced or commensurate, and when the number of dispossessed and displaced persons are proportionally high. Mech­ anisms that should lessen the pains of land dispossession are in many cases nonexistent or too weak. Since other sectors of the economy in many developing countries are usually not capable of providing these people with employment and possibilities for re-adjustment, they are often doomed to join the ranks of unxlix

Introduction

to the second

edition

employed persons or the lumpens (see, for example, Bovin, 1990:37-8, 43 on the Sahel; Salih, 1990:74 on Sudan; Worby, 1988:161 ff. on Botswana). Besides, traditional pastoralists usually lack both the experience and the necessary capital to start market-oriented ranch enterprises. Thus," in Turkey sedentary businessmen and entrepreneurs with managerial experience, not pas­ toralists, have established commercial livestock enterprises. They rent pastures, negotiate transportation, and organize the delivery of stock, meat, and dairy products to the markets. The Yokuk pastoralists have at best been hired as shepherds, but even this is not always the case, since entrepreneurs have often preferred to employ villagers (Bates, 1980:125 ff.; cf. Bovin, 1990:48-50 on the situation in the Sahel; or Little, 1985:145 ff. on the situation in Kenya). Absentee ownership, sometimes ill-hidden, is also becoming more apparent in some parts of the ex-Soviet Union and Mongolia (on the latter see Mearns, 1993). The second major solution or alternative to traditional mobile pastoralism is sedentarization and/or collectivization. Many governments and governmental experts, particularly from Communist and Third World countries, favored and still favor this approach. In the best light, they cast pastoralism as an antiquated and unprofitable economy; in the worst light, they see it as an obstacle in the path of progress. (Some Russian scholars still view pastoral nomadism as an evolutionary dead-end; see, for example, Shnirel'man, 1986:244; Kradin, 1992: 191.) Not infrequently this opinion is inspired by underlying political considera­ tions—a desire to impose upon the pastoralists, or to strengthen the power of central governments. While the colonial powers were often satisfied with the maintenance of general order and taxation, the national governments are in­ terested in exerting direct control over pastoralists. In 1973, when the Sahel was affected by severe drought and many pastoralists there lost their stock and could not migrate anymore, Ebrahim Konate, the Secretary of the Permanent Interstate Committee for Drought Control, expressed his satisfaction with the situation with cynical frankness: "We have to discipline these people, and to control their grazing and their movements. Their liberty is too expensive for us. Their dis­ aster is our opportunity" (Marnham, 1979:9). As an example of an apparently well-intended but nevertheless one-sided and paternalistic attitude, I can quote the principal motion of the Fifteenth Interna­ tional African Seminar held at Ahmadu Bello University in 1979: The conference notes that the nomadic aspect of the life of pastoralists is no longer tenable in the face of ever greater pressure on land, and that it is not in the interest of the pastoralists themselves to continue to lead a nomadic or semi-nomadic way of life. (Adamu and Kirk-Greene, 1986:xvii) Most Sudanese authors, while condemning the failures of government plan­ ning, also continue to praise the ideal of sedentarization. Khognli, 1980 (cf. I

Introduction

to the second

edition

El-Arifi, 1975:157; Salih, 1990.64 ff.), for example, argues that mobility is bad because the time spent walking around represents wasted human capital. The fact that the pastoralists are in no way idle during migrations escaped his attention. However, the results of this policy were usually disastrous both to the pastoralists and to the countries that forced them to sedentarize, particularly if, as it often happened, that effort was conducted by force. In Iran, for example, during the reign of Reza Shah, the government considered pastoral nomadism to be an obstacle to modernization, and even more so to be a politically unde­ sirable phenomenon. It instituted a program of "pacification" and forced the mobile pastoralists to settle, which soon resulted in their impoverishment. About 75 percent of their stock perished, and Iran was deprived of many commodities, such as meat, milk, wool, and draft animals (Irons, 1975; Tapper, 1979:22; Beck, 1986:129 ff.; Black-Michaud, 1986:83 ff.). The Soviet example is even more instructive, particularly because until re­ cently the Soviet propaganda machine presented it to Third World countries as a model of successful development that deserved emulation. The first large-scale experiment of this nature was carried out in the Soviet Union in the late 1920s and in the early 1930s. Within a few years the nomads and pastoralists in Central Asia and other areas were forced to sedentarize and enter newly created collec­ tive farms, which meant giving up their privately owned livestock. Denomadization and collectivization of the nomads met with widespread opposition. Those who resisted were either killed or deported; some people managed to migrate abroad, many died of starvation. The political aims of the center were achieved; the pastoralists were 'domesticated,' but at a heavy cost. Their traditional way of life was ruined and the Soviet economy was seriously damaged. The Kazakhs, in the past the most numerous nomadic people of the USSR, illustrate these tragic events. In the early 1930s, during the traumatic events of forced collectivization and bloody settlement of Kazakh nomads on fixed lands, about 550,000 nomadic and semi-nomadic households were forced to seden­ tarize, often in waterless regions where not only agriculture but pastoralism Population and Livestock, Kazakhstan, 1928-1975 The Kazakh population in Kazakhstan 1930

1939

1959

4,120,000

2,307,000

2,755,000

Number of livestock in Kazakhstan 1928

1933

1975

30,305,900

4,800,600

29,700,000

Source: Khazanov 1990a, li

Introduction

to the second

edition

were impossible. Others were moved to towns and cities to be employed as in­ dustrial workers, but even unskilled work could not be found there (Nurmukhamedov, Savosko, and Suleimenov, 1966:195-6; Olcott, 1987:179-87; Abylgozhin, Kozybaev, and Tatimov, 1989). As these figures show, forced sedentarization cost about two million Kazakh lives and decimated their herds. It took several generations for their livestock to even approximate their pre-collectivization levels. As for the Soviet economy, the source of permanent meat shortages go back to these events in the early 1930s. Those former nomads in the Soviet Union who remained in the pastoral sector of the national economy had to work in the collective and state-owned farms. Alienated from stock ownership, they at best became underpaid wage laborers who were not very interested in the results of their work. It is not surprising that their productivity drastically decreased and their animal husbandry stagnated. In the early 1970s, while doing fieldwork in Kalmykia, a semi-desert region in the lower Volga reaches, I visited one of the best state-owned farms in the area. The stock there was transfered from one pasture to another by trucks, and herds were supplied by water tanks. The manager of this farm, not a Kalmyk but a Ukrainian, was nevertheless a competent professional. The government gave him the highest Soviet award, the Hero of Socialist Labour. He told me about many innovations introduced into the farm's production. He particularly stressed the high productivity of labor that was reached at his farm, pointing out that a herd of 800 sheep needed only 8 shepherds. I remarked that before the collectivization one Kalmyk nomad on a horse assisted only by his juvenile son and a couple of dogs managed quite successfully to tend the same number of animals. At first, the manager did not believe me, but smiling Kalmyk shepherds who witnessed our conversation confirmed my words. Moreover, it was quite obvious that they were rather pleased with them. After a while, the manager uttered his most convincing argument: "OK. I believe you. But such comparisons are wrong. You must take into account that in the past the stock belonged to the pastoralists, and now it belongs to the state." I hardly need to comment on these words. Still, it is remarkable that the general public and many experts in newly in­ dependent Kazakhstan oppose land privatization while advocating private stockownership. Their reason for this is that privatization would inhibit, rather than facilitate, the restoration of the pastoral sector of the national economy. Kazakhs fear that privatization would lead to the appropriation of their lands by nonKazakh people who have the advantage of being more knowledgeable of land markets than Kazakhs. President Nazarbaev of Kazakhstan explains his antip­ athy to outright ownership of land by pointing out that to permit such ownership would be alien to the heritage and mentality of the former nomads (Khazanov, in press). Even when it was not accompanied by sedentarization, the Soviet-style colleelii

Introduction

to the second

edition

tivization was always ruinous for the nomads. Another Communist country, Mongolia, failed in its attempt to obediently imitate the Soviet policy of collec­ tivization undertaken in the early 1930s. In turn, the Mongol nomads revolted and were suppressed only after the Soviet army intervened. Besides, the govern­ ment of Mongolia became aware that it would soon become a government with­ out subjects, who fled abroad en masse. The second collectivization campaign in Mongolia in the 1950s, was accomplished without human losses; however, since that time the pastoralist branch of the economy in Mongolia remains stag­ nant (Bawden, 1989:290 ff.). No wonder that the Mongol government urgently abandoned this system as soon as Communism was finished in Mongolia. Decollectivization and land tenure reforms are still operating in Mongolia and, at best, are in the initial stage in the former Soviet Union. However, many dif­ ficulties lie ahead, especially in forming (or reforming) those institutions that will replace the state ones, such as the existing livestock collective farms. One problem consists in the state bureaucracy's desire to maintain control of land allocation, the sale of livestock products, and the supply of products. It is remarkable, however, that in Mongolia the dissolution of the collective farm is accompanied by the individual households' desire to restore kinshipbased forms of cooperation. During the cultural revolution, the Chinese government also launched a col­ lectivization campaign among its Inner Asian nomads, which greatly damaged pastoralism there. The Kazakh pastoralists were doomed to starvation, and the sons of those who had fled to China from the Soviet Union in the 1930s tried to flee back from the new disaster imposed upon them. However, in the 1980s, the communes were dismantled, stock and pasture plots were distributed among individual families through a contractual system of "household responsibility," and production within the demands of the market was encouraged. At the same time pastoralism became more closely integrated with agriculture. Since then Inner Asian pastoralism is, if not flowering, at least capable of satisfying the local needs and can even produce some surplus (Barfield, 1993:172-6). Fortunately, forced and overtly bloody forms of sedentarization and collec­ tivization have not become a general practice. However, even now sedentariza­ tion sometimes occurs under various forms of external pressures that are put upon the pastoralists, and its occurrence is connected with the prior process of their impoverishment. In such conditions, sedentarization may hardly be con­ sidered a voluntary choice. It is true that the sedentarization as a result of impoverishment of the mobile pastoralists is in no way a new phenomenon; it is almost inseparable from the nomadic way of life. However, in the past the mass and successful sedentariza­ tion usually occurred when the nomads migrated into areas favorable for agriculture and often seized by force arable lands in oases or in zones of dry agriculture (Khazanov, 1992). At present, sedentarization faces additional dif­ ficulties, such as shortages of land suitable for cultivation, demographic presliii

Introduction to the second

edition

sure, etc. It is very difficult to turn to agriculture when arable land is already occupied by other people who highly value it. As a result, former pastoralists often have to sedentarize in marginal areas where agriculture is risky and the results are unpredictable, where they start to cultivate lands that agriculturalists themselves consider of little use for cultivation. Often these lands soon become degraded because of over-cultivation. Moreover, the effective development of even some of these lands depends on expensive irrigation projects and other large-scale capital investments. With rare exceptions, so far neither national governments nor international agencies are in a hurry or able to make such investments. Pastoralists' opportunities in other sectors of the economy are also limited by their low educational level, lack of necessary skills, and high level of unemploy­ ment and underemployment in many developing countries. It is true that the trend to sedentarization may take different forms and be caused by different reasons. In the rich oil-producing Arab countries, where the Bedouin are held in high regard for reasons of social and cultural traditions, the governments are securing their loyalty by investing part of their enormous financial resources in improving the economic and living conditions of their pastoralists. They are do­ ing this through such diversified and sometimes contradictory measures as im­ provements in the infrastructure; school education; health care; land distribu­ tion; job offers, particularly in specific sectors like the military; material sup­ port, including monetary payments; etc. Successful as these measures some­ times may be in preventing the Bedouin from further degradation, they are not always consistently aimed at modernizing the pastoralist economy; moreover, they often reduce the Bedouin's general involvement in the productive sectors of the economy. For example, in the Sultanate of Oman the sharp increase in state revenue from crude oil, after 1973, and the government's desire to stabilize the political situation in the rebellious Dhofar province resulted in developmental programs which included a wide range of material subventions, both in goods and cash, and the creation of thousands of artificial government and military jobs. The pastoralists have become heavily dependent on state support and on earning easy money through the state initiated assistance measures (Jansen, 1986; Moham­ med, 1991; Scholz, 1991; cf. Katakura, 1977; Cole, 1981; Lancaster, 1981; Fabietti, 1982a: 186-97; Fabietti, 1986:22-9; Lancaster, 1986:45-7; Kostiner, 1990:244 ff. on the situation in Saudi Arabia). In Libya, in the 1970s, the countryside of Cyrenaica became depopulated as nomads acquired jobs and positions in the petroleum industry, settled in villages to live in "popular houses" provided by the administration, and took government posts, or collected welfare. Mobile pastoralism began to develop into a marketoriented ranching-type system. This, in itself, could have worked. However, the new trend was based on nepotism and patronage more than on sound economic developments. Those who had good positions invested in herds often tended by liv

Introduction

to the second

edition

Egyptian shepherds, while the remaining Bedouin became dependent on arti­ ficially created jobs in administration or even reliant on welfare (Behnke, 1980). In all such cases one may wonder to what extent the current trend may be con­ sidered a permanent one? What will happen to the former pastoralists in these countries if the government subsidies are withdrawn or reduced due to a fluctua­ tion in oil prices on the world markets, or other reasons? How many stock enter­ prises will go bankrupt without government assistance and heavily subsidized pastoral production? In this respect, what happened in Jordan, which is far from being a rich coun­ try, is quite instructive. Jordan's Bedouin sedentarization was prompted by the provision of various government services and career opportunities in the military and administration. As a result, the few ranches in operation which are owned by the elites cannot satisfy the country's needs, and nearly all of its meat has to be imported (Abu Jaber and Gharaibeh, 1981; Hiatt, 1984). The above brief outline naturally leads one to ask whether it is not premature to dismiss mobile pastoralism as a viable form of economy in many arid areas. It is still worth keeping in mind that mobile pastoralism was originally developed as an alternative to agriculture in just those regions where agriculture was im­ possible or economically less profitable, and in many of these regions the situa­ tion remains basically the same. For example, in many regions of Central Asia herdsmen husbandry, in which herds are pastured all year round, is more profit­ able than cotton production, which was imposed upon the population by the Soviet government and which has greatly damaged both their living conditions and the environment (Khazanov, 1990b; Wolfson, 1990). Bold attempts to drastically change the arid environment have often failed or resulted in disastrous repercussions. For instance, the Russian settlers in Ka­ zakhstan practiced extensive dry farming based on grain monoculture which led to soil exhaustion and diminishing grain yields. In the agricultural development of these lands the American experience with dust bowls was not taken into ac­ count, although the environment and climatic conditions of North Kazakhstan are similar to those of the American Plains with their favorable and unfavorable cycles for grain agriculture. Incorrect agrotechnology applied to the "virgin lands" development brought in weeds and winds, soil and water erosion, and in less than fifteen years generated three million hectares of sand and made another twelve million hectares of land liable to wind erosion (Uteshev and Semenov, 1967:5; McCauley, 1970:100-11; Komarov, 1978:53). A similar fate occurred in the southeastern steppes of European Russia (Kotliakov, Zonn, and Chernychev, 1988:62-3). In Syria, agriculture brought into steppe areas in the 1960s soon failed. While farm machinery was introduced and marginal land was brought under cultiva­ tion, predictable and regular crop production has not materialized. Plowing the steppe had catastrophic effects on the natural environment, including the loss of soil fertility and drastic increases in wind and water erosion, until this practice Iv

Introduction

to the second

edition

was prohibited by law in 1973 (Bahhady, 1981:261 ff.; Lewis, 1987:186 ff.). The survival of mobile pastoralism, and its desirability, are directly connected to the current environmental debates and conservation ethics. It is a remarkable coincidence, even an irony, that after thousands of years of production the pastoralists around the world are suddenly confronting the fruits of their mis­ management in unison. This is, nonetheless, the charge against the nomads nowadays. Across Africa and in other regions they are being held responsible for drought and desertification. The unfounded and fallacious opinion that mobile pastoralism necessarily leads to the deterioration of the natural environ­ ment continues to be reiterated. The emergence and spread of various forms of mobile pastoralism resulted in only one essential ecological change: wild her­ bivores were gradually replaced by domesticated stock, since they competed with the latter for pastures. However, pastoralists had a much more moderate impact on ecosystems than modern ranchers and agriculturalists. Seventy-five million bison were exterminated in the Great Plains in no more than 157 years between 1730 and 1887, but mostly after the conquest of such areas as the Indian Territory and Pine Ridge (Barsh, 1990). In the Eurasian steppes a similar pro­ cess was not completed in three thousand years. It became particularly intense only during the last few hundred years with the expansion of agriculturalists into the steppe zone. Still, herds of gazelles (Gazella subgutturosa) and antelope (Saiga tatarica) survive to the present day in several remote regions of the Eur­ asian steppes where agriculture and industry did not gain an upper hand. Generally, pastoralism does not lead to the deterioration of the floral cover and moderate grazing may even be advantageous to the normal functioning of arid vegetation, since stock trample seeds of wild plants into the soil, eliminate weeds alien to local flora, and fertilize the soil. Under pastoralism, the destruc­ tion of the environment takes place mainly in cases of overgrazing, but vegeta­ tion is able to regenerate as long as overgrazing is temporary, as it is in tradi­ tional forms of mobile pastoralism. So far as the latter depended on a dynamic balance between natural resources, livestock, and the human population, over­ grazing sooner or later led to decreases in the number of people and stock, either through famines and calamities, or through migrations, warfare, and other fac­ tors, which, in turn, eased the pressure on natural resources. Hardin's (1968: 1243-8; Hardin and Baden, 1977) argument about the "tragedy of the commons" has been already disputed by a number of scholars (see, for example, McCay and Acheson, 1987). As a rule, common property on resources does not mean open-access to them (Berkes, et al., 1989:93). In modern times, overgrazing in many parts of the world is connected not so much with cycles of pasture abuse and recovery as with political factors like unconsidered development policy, alienation of the lands belonging to the pastoralists, or a restriction of their movements which increases the pressure on the pastures. In other cases, the over-exploitation of productive ecosystems which gives rise to desertification and degradation of vegetation, soil, and water is caused by im¬ Ivi

Introduction

to the second

edition

posed commercialization and attempts to adapt and apply modern technologies to the pastoralism still operating within the framework of traditional social organization (Reining, 1978). The loss of control by pastoralists over the stock they are herding —the alienation of control of the herds from the immediate pro­ ducers—may also contribute to a serious ecological disequilibrium in places such as the Sahelian zone of West Africa (Maliki, 1986:4-5), or the arid zone of the former Soviet Union. The steppes of Kalmykia and the North Caucasian pied­ mont are rapidly deteriorating now that their seasonal pastures are being used all year round (Kotliakov, Zonn, and Chernychev, 1988:62-3, 69). In Ka­ zakhstan and Turkmenistan, vast areas of fertile pastures have turned into sand deserts due to both overgrazing without a seasonal rotation of pastures and a trend away from multispecies toward monospecies herd composition. Deserti­ fication there is advancing on a scale comparable with the Sahara and the Sahel. In the past the Karakum and Kyzyl-Kum deserts occupied less than 24 percent of Central Asia's and Kazakhstan's total area. By now active desertification has claimed an additional 35-40 percent of the area (Wolfson, 1990:41-2). Attempts to intensify traditional pastoralism and to apply scientific methods to it turned out to be difficult even when the planners did not wish to turn the pastoralists into capitalist ranch owners. On the one hand, in making decisions managers sometimes tend to ignore the benefits of local expertise despite the fact that the traditional technologies have some useful adaptive components. On the other hand, attempts to increase the productivity of pastures and improve breeds, while maintaining the traditional pastoralist way of life and traditional land tenure, often result in undesirable and unexpected consequences, particularly when the problem of balance is ignored. Thus, some African governments advised and assisted by foreign donors and experts launched various water development programs, including the drilling of boreholes to improve pastures and increase the beef production at the expense of traditional dairy-oriented pastoralism. New wells were usually open to every­ body, while in the past the access to water resources was often under the control of certain segments of pastoralists. As a result, pastures around new wells soon became overgrazed (Horowitz, 1981:61-88; Goldschmidt, 1981:104 ff.; Bernus, 1990:166-7). Recent Sahelian history provides us with some sad examples of how hasty attempts at modernization led to overgrazing and, thus, along with droughts, contributed to the deterioration of pastures (Gorse and Steeds, 1987). All in all, although drought and shifting desert boundaries are, to a still un­ known degree, natural occurrences, it is clear that the complex assemblage of modern sociopolitical, economic, technological, and demographic changes are implicated in the progressive decline of marginal arid pasturelands. Well-dig­ ging, ranches, and encroaching agricultural schemes often appear blind to the fact that when pastoralists stay in one place for too long, or run out of adequate space, the natural resources tend to degrade. It seems that, whether it rains or shines, these things will not go away by themselves. Ivii

Introduction

to the second

edition

Ecologically sound management of many arid and semi-arid areas has yet to be achieved, and pastoralism in its mobile forms remains the most suitable form of economy for many regions of the world. Moreover, modernization has often turned out to be painful for the pastoralists themselves. I must also emphasize again that pastoralism is not only a way of making a living; it is also a way of living, dear to those who practice it. The problem, as I see it, is not how to substitute pastoralism with other types of economic activities, but rather how to make it more efficient and how to make the inevitable transformation of the traditional economy and social organization of pastoralists less painful to those involved. There is no single recipe which applies to all countries and all groups of pastoralists the world over; however, it should be stressed that the pastoralists are quite capable of appreciating those innovations which they consider bene­ ficial. The spread of trucks amongst the Bedouin of Jordan, Syria, Iraq, and Saudi Arabia (Chatty, 1986) is only one indication of this. One thing, however, seems clear enough. Excessive paternalism, even a ben­ evolent one, will hardly help. We will not have economic development until we permit, stimulate, and encourage self-development. Planners come, planners go, even if they do more than just plan in their armchairs; subsidies and support are provided and withdrawn; governmental policies and priorities change. But the pastoralists themselves remain facing their harsh natural environments. It is they who have to pay the price for ill-devised projects and all those decisions made by others which are often based on purely political considerations ignoring the pastoralists' interests. Only when the views of the pastoralists themselves are taken into account, and their own participation in the decision-making process is secured, can we begin to expect their successful modernization. I must warn the reader about some inconsistencies in geographic terminology used in this book. In the first edition I have followed the old Russian tradition and called "Middle Asia" the region which in the Western tradition is usually called "Central Asia." However, the newly independent states in the former Soviet Middle Asia, apparently for political reasons, have declared their desire that their region be called "Central Asia," and nowadays this term is beginning to be used even in Moscow. Since I cannot, nor wish to, withstand this trend, I obey. Although I did not change the terminology in the text of the book, in this new introduction "Central Asia" stands for "Middle Asia." I am sorry for this incon­ venience, but the ongoing political process in the region is beyond my control. I thank the publishers of Cambridge University Press for transferring the copyrights of this book to me and the publishers of the University of Nebraska Press for permission to reproduce in this new introduction some materials of my article "Pastoral Nomads in the Past, Present, and Future: A Comparative View," first published in The Struggle for the Land, Paul A. Olsen (ed.), Lincoln and London, University of Nebraska Press, 1990:81-99.

Iviii

Introduction

to the second

edition

During the years of my living in the West I have had many occasions to meet with my colleagues and discuss many anthropological and historical problems connected with pastoralists. I greatly benefited from my contacts and conver­ sations with Robert McC. Adams, Thomas Allsen, Victor Azarya, Ofer BarYosef, Thomas Barfield, Clifford Bosworth, Michael Chamberlaine, Patricia Crone, John Davis, Shmuel Eisenstadt, Ugo Fabietti, Ernest Gellner, Yosef Ginat, Peter Golden, Stephen Hodkinson, Tim Ingold, John Karras, Mark Kenoyer, Thomas Levy, Herbert Lewis, Rudi Lindner, Beatrice Manz, Emanuel Marx, Robin Mearns, Thomas Noonan, Steven Rosen, Uli Schamiloglu, Haim Tadmor, Arthur Waldron, Andre Wink, and many others. None, of course, bears any responsibility for my possible errors and omissions. I am also very grateful to my friends, informants and colleagues in different parts of the former Soviet Union 'who continue to provide me with publications unavailable in the West, as well as with information on the recent status of the pastoralist peoples in their countries. My graduate student Jeff Kaufmann was a great help to me in the preparation of this new introduction. My particular gratitude goes to the students who attended my classes and seminars on pastoralism at the Hebrew University in Jerusalem and the Univer­ sity of Wisconsin-Madison. Their curiosity, interest, often fresh outlooks and criticism of the views that the older anthropologists, including myself, some­ times tend to take for granted, are always a source of encouragement and stimulation to me.

lix

Introduction The phenomenon of nomadism: myths and problems

Poets of different times and of different ages have written many poems (whether or not they are any good is a matter of taste and of the quality of the poems themselves), exalting the beauty of the steppe or, correspondingly, of the desert or the tundra, and the delights of nomadic life. Strictly speaking there is nothing particularly remarkable in this. Indeed, for many traditional societies there is nothing more familiar and therefore already more beautiful than the space surrounding them. But what is far more curious is that in this category of poets there are many non-nomads alongside nomads. The attitude of sedentaries to nomads has always been ambiguous. The myth of the nomad may be even older than the myth of the 'noble savage'. By the middle of the fifth century B.C. in the writing of the father of history, Herodotus, idealized descriptions of nomad Scythians may be found. Further and even more idealized descriptions may be found in Ephorus. In the fourth century B.C. the Stoics and Cynics seized upon them in order to contrast natural and unspoilt barbarian life with the vices of civilization. The myth of the nomad even carried over into the Middle Ages. Possibly it played a part in the development of the legend about the eastern kingdom of Prester John, who was called on to free the Christians in the East from the yoke of the infidel. In modern times both myths (of the savage and of the nomad) have been revitalized, but the myth of the nomad would seem to be the more lasting one. Philosophers have influenced travellers and travellers have influenced scholars. Many sources and ideas have misinformed poets, writers and the general reading public; at the same time there has been a great desire and search for such misinformation about nomadism. In its striking noncon­ formity with the sedentary life of townsmen, the image of nomadic life has exercised the strong attraction of opposites. This negative approach to reality has resulted in different outlets, some in the 'realm of fancy'. A stereotyped view of nomads has arisen in which their real or imaginary freedom and political independence almost occupy pride of place. Moreover, despite its poverty and other drawbacks, nomadic life is I

Introduction thought by nomads themselves and by many onlookers to have one important advantage, which was defined by A.C. Pigou at the beginning of the century as 'quality of life'. Myths last for as long as there is some need for them. At the same time stereotypes change with difficulty. It could be said that the Hollywood ideal of the cowboy is a direct or indirect descendant of the myth of the nomad. At least they can essentially be seen as one and the same thing. But myth has a purpose. It is myth which keeps us from knowing the half tones. Apart from its light side the myth of the nomad also has a darker side in which the nomad is perceived almost as the devil incarnate. In China the Confucians simply despised the barbarian nomads, considering them incapable of following a civilized way of life. But in the West, both before and after Jordanes, who believed the Huns were descendants of evil spirits and witches (Getica, 121), there were many people who had the same idea. From the times of the biblical prophets the view that nomads are excessively savage and wild has often been associated with the view that they have a particular destiny, as a means through which God can chastise different peoples. The early Christian authors were familiar with this idea. Not much later Attila was described as no less than the 'scourge of God'. In the Middle Ages similar notions were very widespread and popular; nomads were often included in the register of disasters which listed, amongst others, cholera and the plague. Once they were no longer associated with disasters, nomads were seen in a more positive, although still exotic light. As it turned out, the dark side of the myth was less lasting than the light. However, this did not mean that the dark side of the myth disappeared altogether. It has been used more than once in colonial and even in post-colonial times by politicians as a means with which to manipulate public opinion. The realm of fancy and the sometimes falsified, or at any rate one-sided, depiction of earthly reality often have turned into neighbours. The time has come for us to ignore those myths. If nomadism is to be approached as the final result of specialized pastoral economies, and the question posed as to whether nomadism represents more in the evolution and history of mankind than one example of economic adaptation, then we must at once distance ourselves from pressure generated by the selfappraisal of nomads themselves and also from the subjective and emotional evaluations of other, very different societies, in which there are other forms of adaptation and other evolutionary alternatives. Is it permissible to look on the phenomenon of nomadism as something different from the phenomenon of the sedentary sea-mammal hunters of the Arctic Ocean, or the sedentary fishermen in the deltas of certain rivers on the northwest shores of North America, or even of the different types of

The phenomenon

of nomadism:

myths and

problems

shifting horticulturalists in the tropics? Are we not here merely dealing with one level of economic development (a food-producing rather than a food-extracting economy), or with the specificity of the given form of economy (in the final analysis, all forms are specific), or just with something which is quite well known, but not closely connected with academic definitions? In the final analysis a phenomenon is not only a rare, unusual and unique occurrence, it also may be widespread and, most importantly, global in its consequences. The Tasmanians were a rare phenomenon; the Bushmen of today are still a rare phenomenon. But in my opinion the phenomenon of nomadism also consists in quite the opposite: (a) not only as a specialized occurrence, but also as a very widespread one (particularly in the past), existing all over the world, except in Australia and to a certain degree in America; (b) in its role of linking different societies and cultures; (c) and finally, not only in its economic, but also in its social and historical specificity. A paradox arises which anticipates the content and conclusions of the book and which may be formulated briefly in the following way: societies based on one of the most specialized types of food-producing economy, in which technology is relatively conservative and has changed little with time, have exercised an essential and, indeed, multifarious influence on the social and political functioning and evolution of non-nomadic societies in which the economy is more diversified and technology more advanced. Amongst nomads themselves pastoral specialization has meant more or less economic one-sidedness and no autarky, and outside the society proper these have led directly to social mobility and heightened political activity. Most importantly, nomads could never exist on their own without the outside world and its non-nomadic societies, with their different economic systems. Indeed, a nomadic society could only function while the outside world not only existed but also allowed for those reactions from it - reactions which were economic, social, political, cultural, in a word, a multi-faceted response - which ensured that the nomads remained nomads. In this way, in my view, the important phenomenon of nomadism (while it remains nomadism) really consists in its indissoluble and necessary connection with the outside world; that is to say, with societies which have different economic and social systems. This book is devoted to the establishing of this thesis. In the humanities and in anthropology, the reduction of the complex to the simple and of many factors to one often makes more accessible the thought-process of the scholar and the result of his research, but rarely does it truly bear fruit. That which in the natural sciences is possible, or at least considered to be ban ton, is not usually applicable to the humanities. In this respect the problem of nomadism serves as a graphic example. 3

Introduction It is not in any respect a closed system. Hence, attempts at defining the operation and evolution of nomadic societies only from the inside, from the viewpoint of ecological or socioeconomic determinism by their own environment, can never be fully successful. Admittedly, such attempts are usually no longer avowed, because the idea of determinism is no longer fashionable; but in practice they are still not infrequently encountered. The detailed research of many scholars proves that while ecology and economy are certainly amongst the most important factors (although they are not the only ones) which determine the specific character of nomadic societies, these factors still leave a great deal of scope for other variations. Apart from the economic there are also sociopolitical ones. Most importantly, these variations are connected not only with intrinsic factors, but also with extrinsic ones. The attempts which have been made up to the present day to extend unconditionally the different global and universal systems of the historical development of mankind to include nomads seem unjustifiable. My concern at this point is not to verify or to assess the usefulness of one or other such system; it is a rather different one. Those systems were constructed around data concerning the evolution of sedentary societies (often" only in certain defined regions) and mechanically included nomadism, taking no account of the specificity of the latter. However, the connection between nomads and the outside world is one thing; the complete identification of nomads with the outside world is another. To underestimate the idiosyncracies of nomadic societies is just as dangerous as it is to overestimate them. The problem of 'nomads and the outside world' is, therefore, so complex and many-sided that there seems to be no way in which it can be simplified. That which is true of one given nomadic society must be reassessed when applied to another nomadic society. Historical perspective must be maintained. That which is true of nomads today may not apply to the nomads of ancient time or of the Middle Ages. No one scholar could possibly compile something approaching a complete and qualified description of a number, let alone of all nomadic societies; a diachronic analysis, as opposed to a synchronic one, would be even more impossible. That time has passed in which one scholar-compiler may attempt a compendium of the Weltgeschichte type. One way round the problem is to try to find typologies, models and generalizations, i.e. an inevitable and conscious simplification and schematization of reality. If the author is to follow this path he must be prepared for a sceptical reaction from his colleagues; he must realise that there are many superficialities in his research and many very debatable points, possible inaccuracies and even mistakes in his work which, at opportune moments, his colleagues will readily and kindly point out to him. In a word, the author 4

The phenomenon

of nomadism:

myths and

problems

must be prepared not for dithyrambs but for some basic and severe criticism. This author has decided to undertake the task of assessing whether or not there do exist some definite patterns in the interrelations between nomads and the outside world (rather than making another study of specific nomadic societies or the nomads of one particular region), for the precise reason that he feels the time has come for such an assessment to be made. It is now as necessary to see the wood in its entirety behind the individual trees as it is to see the individual trees in the wood. The geographical terminology used in this book may not be familiar to everybody. For reasons which will become clear below the author differentiates between the Near East (Iraq, the Arabian peninsula, the countries of the Fertile Crescent, North Africa down to the Sudan and the Horn of Africa) and the Middle East (Turkey, Iran and Afghanistan). The term Middle East entered common use for rather circumstantial reasons which are connected with the Second World War. As Fisher (1966: 1-2) asserts, 'Up to 1939, there prevailed a somewhat vague and loose division of southern Asia into Near, Middle and Far East . . . The war of 1939 at one stroke removed the question of territorial definition in Western Asia from the academic groves to which it had hitherto been mainly confined. There came the fait accompli by which a military province stretching from Iran to Tripolitania was created and named "Middle East".' An anthropologist need scarcely be restricted to such a use. In accordance with established academic traditions in Russia and partly in continental Europe, in this book Middle Asia is differentiated from Inner (Central) Asia; in the Anglo-Saxon tradition both regions are referred to under the general term, Central Asia. Historically the regions are very closely connected. However, anyone who has visited them or who is acquainted with literature about the geography of the area knows that there are important environmental differences between them. Consequently, for this author Middle Asia comprises the region which is flanked in the north by the Aral Sea and the Kazakh steppes, in the south by the Kopet-Dagh and Hindu-Kush, in the west by the Caspian Sea, and in the east by the Pamirs. Inner Asia is that region which comprises Kashgaria, Jungaria, Mongolia and Tibet, and is flanked in the south by the Himalayas and in the northeast by the peaks of the Inshan and Great Khingan. In this book more space is devoted to materials about the nomads of the Eurasian steppes, semi-deserts and deserts than to materials about nomads of other regions. This is for no other particular reason than that it just so happens that most of the work about nomads and nomadism which the author has published in the past has centred on this region. Both from his reading of relevant literature and his fieldwork - ethnographical in Middle Asia, Kalmuckia and Daghestan, and archaeological in the South Russian steppes, Middle Asia and tthe North Caucasus - he knows Eurasian nomads 5

Introduction and nomadism more thoroughly than he does the nomads of any other region. An entire chapter of this book is devoted to the origins of pastoral nomadism. This is not only because the author is not satisfied with the current hypotheses and theories. He also considers that in a book about the interrelations between nomads and the sedentary world it would seem to be both interesting and expedient to begin the history of the interrelations ab ovo. I cannot but agree with Boas (1940:305): 'To understand a phenomenon we have not only to know what it is, but also how it came into being.' It is partly for this reason that particular attention is paid wherever possible in this book to nomads of ancient times and of the Middle Ages, and partly also for reasons of a different kind. The nomads of today directly or indirectly already have been drawn into contemporary economic and political systems of the sedentary world. However, as a rule, this has happened by different means and on a different basis from the way it did in the past. The character of the interrelations between nomads and the outside world has undergone qualitative changes. Changes now occur in nomadic societies so quickly that anthropologists often do not have time to record them. Thus, where the present century is referred to in this book the use of the present tense is often purely conditional. Of course the modern period did not begin everywhere at the same time. But the origin of the process which fundamentally changed the position of nomads in the world was already determined in the late Middle Ages by the emergence of strong centralized states, such as the Ottoman, Russian and Ch'ing empires. The colonial period enabled this process to intensify and become more widespread; and in post-colonial times corresponding tendencies only increased and became more firmly established. The question as to whether nomadism can survive in the contemporary world raises serious doubts. Opportunities for nomadism to adapt itself to the outside world are few and far between. Once again Cain is killing Abel, slowly but surely and with very little standing in the way, this time insisting on the most noble of intentions. It is true that today, by way of exception, the short-lived revival of nomadism in certain areas may be observed. But the revival is local in character and scarcely will last for long. Consequently, in many instances it is risky to extrapolate about the past from contemporary field work and its conclusions without duly considering the changes in the character of the interrelations between nomads and the outside world and their consequences, which have taken place over the course of the last two or three hundred years. But the data about the nomads of ancient times and the Middle Ages have one essential defect. As a whole facts are relatively few and far between and, moreover, they were collected not by professional anthropologists, but by 6

The phenomenon

of nomadism:

myths and

problems

people who at best were well acquainted with nomads and their way of life but who, on the other hand, had not read Lattimore, Barth or Dyson-Hudson. Nevertheless, any attempt to trace the particularities of nomadism not only in its present rather sad state, but also in a historical context, must take these facts into consideration. The position is made somewhat easier when reference is to the ecological and productive fundamentals of nomadism, which are least well documented in the historical sources but which, on the other hand, have changed relatively little with the passing of time; in this sphere extrapolation from contemporary facts or facts from the not too distant past is not only inevitable, but even permissible within certain limits, although many reservations must also be acknowledged. But when reference is to the sociopolitical organization of nomads and, particularly, to concrete forms of their interrelations with sedentary peoples, reference to historical sources is imperative. Fortunately, since ancient times the sedentary world has not only been concerned with mythopoesis when dealing with nomads. Discoveries are not made in the course of general observation, but at the time when it becomes the common interest of mankind for them to be made. Nomads, it would appear, have never taken any initiative on this account. Their knowledge of the outside world has remained for the most part utilitarian. On the other hand, from the time of its first acquaintance with nomads the outside world has been interested in different aspects of the lives of nomads and this interest, at first utilitarian and emotional, gradually became more academic. The detailed research of that theme is the subject of another book, which should be very useful and interesting, but it has yet to be written. Here I shall only briefly mention the basic stages in the development of the study of nomads which is far from complete, if indeed there is anything which is complete in this world. Even before the emergence of real nomadism the ancient states of Western Asia were forced to have dealings with mobile pastoralists and semi-nomads. Descriptions of their way of life stress how unusual they seemed to sedentaries, but there is no special ethnographical information in these descriptions. In the first centuries of the last millennium B.C. the pastoralists and seminomads of the previous millennium B.C. began to be replaced in certain cul­ tural areas of Western Asia by real nomads who often were migrants from other areas. Those were times of shock and terror, caused by the unwel­ come appearance of a menacing and hitherto unknown force. The biblical prophets give what is perhaps the best description of all this (see, for ex­ ample, Jeremiah 5.15 17). In the Bible and in Assyro-Babylonian sources 7

Introduction some facts are to be found about the political history of nomads in this period, but there is no purely ethnographical information. The first stage in the study of nomadism comprises antiquity and part of the Middle Ages in which the ancient tradition was continued and preserved. In this period nomads become a constant, although usually a disturbing and often disagreeable factor in world order. Administrators and military commanders in sedentary lands who had to have dealings with nomads rarely bothered with academic research. Scholars for the most part lived far removed from nomads and, with only individual exceptions, were usually content with standard descriptions of the way of life and mores of nomads, which were so different from the way of life and mores of agriculturalists and townsmen. These descriptions report the absence of agriculture, movements with livestock according to availability of grass and water, and the absence of fixed movement patterns; they are repeated sometimes almost word for word in Greek, Roman, Chinese and even in some medieval sources, and often appear to be verbatim copies of each other (see Herodotus IV.46 on the Scythians, or Diodorus IX.94.2 on the Nabataeans, or Ssu-Ma Ch'ien [Syma Tsian], Shih Chi, Ch. 10 on the Hsiung-nu, or Ammianus Marcellinus XIV.4.3.5 on the Saracens and XXXI.2.10 on the Huns). Nevertheless, the ancient tradition provides us with an idea of a particular nomadic world within which all the individual peoples are similar, but at the same time quite dissimilar from other peoples living outside the confines of that particular world. This idea, reinforced by concrete empirical ex­ perience as well as by speculation, has turned out to be very long-lasting. Scholars in the Caliphate succeeded in widely disseminating the ideas of their predecessors in antiquity. According to the twelfth-century author Marvazi who, with reference to Hippocrates (now identified as PseudoHippocrates), wrote about the nomads of the east-European steppes, the Scythians of the distant past had become the Turkic peoples of his day; at least in no essential way were the two peoples different, for the same author goes on to insist that'. . .in Europe all the Turkic peoples are alike, but they are like no other people'. For the Greeks in antiquity the word 'Scythian' was associated with a specific ethnicity and generally with the nomads of the North; the word 'Turkic' meant much the same for scholars in the Caliphate. Such views were reinforced by ideas about the nomads' particular type of economy and its dependence on natural environment. The beginnings of the ecological trend in the study of nomads are thus to be found in antiquity. At the same time it was during this early stage that the first isolated attempts to include nomads in the general schemata of the development of mankind were made; true, these mythologized about a lost golden age or the antithesis of this, or else they can be found in rather more scholarly form in 8

The phenomenon

of nomadism:

myths and

problems

the tripartite theory worked out by Varro. Finally, the wisest and most far-sighted of the more pragmatic politicians, whose aims in examining the social and political peculiarities of nomads were eminently practical, concluded that in these peculiarities were contained both the strength and weakness of nomadism. Maybe because of pragmatic necessity, therefore, this stage witnesses the beginnings of the sociopolitical trend in the study of nomads. The second stage begins in the Middle Ages and includes part of the modern period. Partly it continues the traditions of the preceding period, extending and enriching them, partly it witnesses much which was new in the study of nomads, preparing the way for the following period. Of course the emotions and impressions of many of those who personally lived through or witnessed the invasions of nomads scarcely differ from the feelings of the biblical prophets as they describe the descent of the Cimerians and Scythians On Western Asia. But Rashid al-Din and Ibn Khaldun were not ruled by their emotions. As politicians they recognized that nomads were an immutable part of political life. They were also scholars. Thus, they are worthy of special credit. In their works the connection between nomadism and a specific natural environ­ ment is for the first time given serious consideration; they make considerable progress in the study of the social organization of nomads. Rashid al-Din was one of the first to pay serious attention to the meaning of kinship and genealogical ties for nomads. And finally, it is in the works of both these scholars (and also of their less well-known precursors and followers) that nomads for the first time become not a by-product or antipode of civilization, an abortive line of history, but an integral, constituent and even vital part of history. The whole cyclical conception of civilization devised by Ibn Khaldun would have been impossible without nomads. The third stage in the study of nomads is connected with the establishing of modern history and anthropology. In Western Europe the horrors of nomad invasions were increasingly becoming something of the past, and so the study of nomadism here became less empirical than it was in the East. In the West attempts were renewed, or rather they began afresh, to define in general terms the place of nomads and pastoralists in evolution and history. The tripartite theory was extremely popular in the eighteenth and nineteenth centuries; but in the minds of many scholars it left open many of the same questions which had troubled Ibn Khaldun. Many of those scholars tried not only to define, but also to evaluate the role of nomads in the general development of mankind, and in the origins of the state and of civilization. At the same period as Montesquieu and certain other scholars were trying to establish the egalitarian character of nomadic societies, Fergusson and

Introduction Adam Smith were claiming that the development of pastoralism led to social differentiation. Kant saw the origins of statehood in the conflict between nomads and agriculturalists. Hegel, on the other hand, linked nomads with the second prehistoric period which preceded the emergence of the state, even though it is characterized by conflicts. In this way, towards the beginning of the nineteenth century, a specific tendency was formulated, or perhaps it would be better described as a state of inertia in the theoretical study of nomads; nomads and pastoralists (no precise differentiation was made between the two) were looked upon not so much as a particular ethnographical phenomenon, but as a factor in the general development of mankind. Nomads were being characterized not only by sedentaries, but by sedentaries who were civilized townsmen. Nomads were basically of interest to scholars in so far as they could be fitted into different historical conceptions and schemata, which were based on the agricultural and urban development of society and crowned by civilizations created by agriculturalists and townsmen. It is true that in the eighteenth century, and particularly in the nineteenth, yet another side of the study of nomads, which for a long time had had only a tenuous connection with the one mentioned above, rapidly began to emerge. Philosophers and historians remained indoors, confined to their studies, to argue about nomadism, but field researchers were out at work in Africa, the Near and Middle East, the Eurasian steppes and Inner Asia. These researchers were not, however, professional anthropologists and were often moved by purely practical necessities, but they were responsible for the first specific descriptions of specific nomadic peoples. These descriptions were often not very skilled, even less were they specialized, but what they did was to provide a quantity of much needed primary material which was ready for reworking. The fourth stage in the study of nomads can be dated to the first half of the twentieth century. Characteristic of this stage, rather surprisingly, was the fact that, on the one hand, the study of nomadism became much more anthropological and, on the other hand, a distinct gap formed between the development of anthropological theory and its application to nomadic societies, despite the fact that field research had begun to be undertaken by professionals. This state of affairs is noted by N. Dyson-Hudson (1972) who offers an explanation for it. Admittedly, in general theories of the Kulturkreislehre type, materials relating to nomads are allotted a great deal of space, but they are only used as examples which corroborate theoretical positions already formulated. Admittedly, such scholars of the end of the last century and the beginning of the present one as Ratzel, Gumplowicz, Oppenheimer and Thurnwald showed a theoretical interest in nomads, but only as an active factor in I he 'conquest theory' of the origin of the state which they were developing at the 10

The phenomenon

of nomadism:

myths and

problems

time; with this theory they were carrying on the old line of research about nomads. At the same time, however, there were new theories and new theoretical approaches: areal, socio-cultural and environmental adaptation, functionalstructural, cross-cultural, psychological, neo-evolutionistic and ecological. They were many and varied, but made comparatively little use of data about nomads, and in turn did not have any great influence on the research into nomadic societies. The gap between historical and anthropological research in the field of nomadism has not narrowed, rather it would seem to have become even more pronounced. In the many concepts of the evolution and history of mankind which have been put forward by different scholars, adhering to different schools of thought, with the exception of Toynbee's works, nowhere are nomads apportioned a fitting, let alone, a special place. At the same time anthropologists have concentrated on synchronic research and rarely resorted to using historical materials. In every respect Lattimore's brilliant book (1940) stands out on its own. In the Soviet Union the study of nomadism has been carried out within the framework of the Marxist line of scientific thought as formulated in the thirties and forties. Characteristic of this period are the predominance of speculative calculations over concrete research, of theoretical office-work over fieldwork, and the desire to prove that the development of nomadic societies is in no essential way different from the development of sedentary societies, for both correspond to historical materialism's accepted postu­ lates about the universal socioeconomic stages which consistently succeed each other. It is hardly surprising that it was in this period that the slave-owning and feudal stages among nomads were 'discovered'. On the other hand, because the Soviet scholars working in those years were under the indubitable influence of Marxism and also because their academic tradition was different to the one which existed in the West, there was no discrepancy between historical, archaeological and anthropological theories in their research. The fifth stage in the study of nomadism begins in the West and in the Soviet Union at roughly the same period (somewhere in the mid fifties) and continues up to the present day. However, it is possible that a new stage began in the mid seventies. Whether or not this is the case only the future will tell. Characteristic of this fifth stage is a marked expansion of field research into nomads and an increase in the number of scholars working on different problems connected with nomads and nomadism. Materials about nomads have caused many different spheres of contemporary anthropological thought to begin to be adapted and scrutinized thoroughly. In the Soviet Union political changes in the post-Stalin period have been reflected in the 11

Introduction academic world; thus in Soviet anthropology a great deal more attention has begun to be paid to the ways in which nomadic societies differ from sedentary ones, to the particularities of their sociopolitical organization and to the specific ways in which they function and develop. Discussion of these problems began in the fifties and continues in a diverse form up to the present day. In recent years the ecological approach to the study of nomads to some degree has been extended. When referring to the particularities of the fifth stage I have deliberately avoided mentioning the names of scholars who are connected with specific theoretical approaches and paradigms. It is extremely difficult when referring to colleagues who are still alive, still working actively and fruitfully, to avoid subjective evaluations. Consequently, it is best that the description of their theoretical views either be left to them themselves, or to historiographers of the future. I have also deliberately chosen to avoid at this point a discussion of the comparative merits and shortcomings of specific works about nomads which have been published recently. I am not doing this out of tact or fear of offending one or other of my colleagues. It is necessary that at certain points later on in the book I evaluate certain works, some of them critically. But to criticize is easier than it is to create. Thus, I should like to stress in this Introduction that it is my strong conviction that all the scholarly works which I have read about nomadism, old and new alike, are both valuable and useful, and that all academic approaches in research into the phenomenon of nomadism have their advantages. This in no way entails whole­ hearted agreement with them; freedom from prejudice is all that is required. At the same time, however, I should like to stress yet another factor. Now that a sizeable amount of material has been accumulated and theoretical research has made significant progress, nomadism once again must be studied in all its complexity, not only in its static but also in its dynamic functioning. It must be studied as an aggregate whole with its inner and outer systems, structures, functions and ties, not only as an isolated variant of local, specific or multilineal evolution, but also as an essential and integral factor in human history. Finally, it is necessary to put an end to the almost complete predominance of synchronic research and to return to diachrony its proper place in the anthropology of nomadism. With reference to nomads I can only repeat the incisive conclusion of Evans-Pritchard, reached in the fifties:'. . .by and by anthropology will have the choice between being history and being nothing'. Of course, work with historical sources (as, incidentally, with any others) is linked with specific difficulties; of course, these sources are not complete, at times they are imprecise. Nevertheless, much useful and indispensable anthropological information can be extracted from them. A palaeontologist 12

The phenomenon

of nomadism:

myths and

problems

will not turn up his nose at a tiny piece of bone, in the same way as an archaeologist will not do so at the handle of a broken pot; they are satisfied with what they can find. In order to become a palaeoanthropologist, an anthropologist must reconcile himself to the fact that he must make do with available historical material and learn to use it in the best possible way. In comparison to historical criticism, anthro­ pological criticism of written sources is as yet in a state of infancy. However, in my view, the latter contains reasonable possibilities and perspectives. The belittling of reports made by early travellers and observers of nomadism is quite unjustified. Of course, it is a pity that neither Volney or Burckhardt read Emmanuel Marx or Bonte, although there are still advantages in this fact. But does this mean that their reports are then of no value to the anthropologist? Considering the changes in nomadic societies which have occurred since the times when their accounts were published, anthropologists are faced with two alternatives: either they can confine themselves to a study of the situation in the present day; or else they can use the materials of their precursors and, fully aware of contemporary scientific knowledge and methods, they can try to extract the maximum amount of useful information from these materials, rather than just point to their obvious inadequacies. The union, or at least the dialogue, between anthropology and history is perfectly practicable and it promises reasonable fruit. There are precedents for this already, for example, in classical studies. In order to avoid digression here it will be sufficient to list such English scholars of past generations as Andrew Lang, Jane Harrison, Gilbert Murray, H.M. Chadwick and, moreover, our contemporary, M.I. Finley. If history up to now has gained more from this union than anthropology, this is only because historians began to use anthropological materials and anthropological methods of research before anthropologists began to use historical ones. Finley (1975:108) is correct when he writes, '. . . beginning with Malinowski, anthropologists over-reacted to the historical conjecturing and the unilinear evolutionism of their predecessors by rejecting not only their bad methods but the subject of their inquiry as well, a procedure which, though understandable, is not justifiable'. I can only add to this that the not uncommon identification of Marxism or unilinear evolutionism with historical explanation in anthropology is scarcely justified. The historical approach is entirely compatible with the different anthropological schools and paradigms. The phenomenon of nomadism once again needs to be looked at as a whole, with both the research of the past and contemporary academic conceptions being taken into account. The ideal situation would be to see once again the whole wood in its entirety, without at the same time letting 13

Introduction any one tree out of sight. This task is too much for one scholar, but this author is glad that he is not alone in his aspirations. Efforts in this direction are now being made and, it is to be hoped, they will continue to be made in the future. Surely, in the end, de nihilo nihil.

14

1 Nomadism as a distinct form of food-producing economy

What is nomadism? Terminology is something about which we should agree, not argue. Unfortunately, and in this respect nomads are no exception, this rule is often not observed in practice. The term 'nomads' means different things to different scholars; however, for a long time now it has been used to describe two basic tendencies. On the one hand, some scholars have defined nomads as all those leading a mobile way of life independent of its economic specificity; other scholars have described nomads as extensive and mobile pastoralists who either have nothing at all to do with agriculture, or who are occupied with agriculture to a limited degree in the capacity of a secondary and supplementary activity. The first notion of nomadism, which had gained too wide a currency, appears gradually to be going out of use; however, it may still be encountered (e.g., de Planhol, 1966:277 calls the Australian Aborigines nomads; Lee and De Vore, 1968:11-12 call all hunters and gatherers in general nomads; while Averkieva, 1970:3f., uses the term for mounted hunters of bison; see also Murdock and Wilson, 1972:256-7; Simonsen, 1972:190). In my view, wandering hunters and gatherers, on the one hand, and mobile pastoralists, on the other, have too little in common to unite them under a single label. The bases of their economy, food-extracting in the first instance, food-producing in the second, are different in principle; thus their reasons for being mobile are different and the character of the mobility is different. Kroeber (1947) obviously had this in mind when he contrasted 'pastoral nomadic' with 'primitive nomadic'. In the same way, the term 'nomads' is not applicable to other mobile groups, whether ethnicprofessional groups such as gypsies, or the so-called 'maritime nomads' of Southeast Asia, or shifting horticulturalists, or certain groups of workers in contemporary industrial societies (so-called industrial mobility). Consequently, hunters and gatherers who do not lead a sedentary way of 15

Nomadism

as a distinct form of food-producing

economy

life are best described by the term 'wandering' (correspondingly, semiwandering, semi-sedentary, etc.), and mobile extensive pastoralists by the term 'nomadic'. Recently a similar view has been held by many scholars (see, for e.g., Forde, 1963:33-4, 406; von Wissman and von Kussmaul, 1959:874; Krader, 1966:408-9; Spooner, 1973:3; Khazanov, 1975:5-6; Andrianov, 1978:120). In this way we are returning to the original meaning of the term and the sense which the ancient Greeks attributed to the words 'vouac;, vouxtSec;, vo(xa6ixog'. 'And the wind returneth again according to his circuits [Eccl. 1.6].' However, if all mobile pastoralists are described as nomads this once again leads to an excessively broad and imprecise use of the term, because there are very many different forms of mobile pastoralism. It is obvious that the definition of pastoral nomadism as a particular form of food-producing economy should be based on the sum total of those economic particularities in which it differs from other kinds, forms and even varieties of economic activity. In my view, the most important characteristics defining the economic essence of pastoral nomadism are: (1) Pastoralism is the predominant form of economic activity. (2) Its extensive character connected with the maintenance of herds all year round on a system of free-range grazing without stables. (3) Periodic mobility in accordance with the demands of pastoral economy within the boundaries of specific grazing territories, or between these territories (as opposed to migrations). (4) The participation in pastoral mobility of all or the majority of the population (as opposed, for example, to the management of herds on distant pastures by specialist herdsmen, into which only a minority is involved in pastoral migrations). (5) The orientation of production towards the requirements of subsistence (as opposed to the capitalist ranch or dairy farming of today). This fifth characteristic today no longer applies, or applies only in part to certain groups of pastoral nomads which have been drawn into the world market system; but it was fairly characteristic of pastoral nomadism in the past, although even then pastoral nomadism was not economically self-sufficient. Production was not at that time aimed at specific profits, although often it was directed to quite a considerable extent towards exchange. 1

These are the basic economic characteristics of nomadism. On the one hand, they underline its specificity as a distinct form of food-producing economy, thereby distinguishing nomadism from other kinds and forms of economic activity and indicating its limitations; on the other hand, they permit us to unite under one heading and systematize similar types and 1

It is necessary to differentiate between regular pastoral migrations (for example, the mobility of the nomads of the Eurasian steppes to seasonal pastures) and irregular ones (for example, amongst the nomads of the Sahara); but in pastoral nomadism sooner or tnlci mobility is inevitable.

16

The basic forms of

pastoralism

sub-types of economic activity (in the given instance of pastoral nomadism), over and above other secondary characteristics in which they may differ. Finally, with those characteristics, to a greater or lesser extent, are connected the basic economic, social, political and cultural parameters of nomadism, and even its interrelations with the outside world. In this way pastoral nomadism may be defined from the economic point of view as a distinct form of food-producing economy in which extensive mobile pastoralism is the predominant activity, and in which the majority of the population is drawn into periodic pastoral migrations. Perhaps one can also add that pastoral nomadism, in common with the other main forms of economy, is associated with a particular level in the development of technology. This level characterizes the period between two revolutions (including their consequences): the Neolithic and the Industrial. According to the suggested definition, pastoral nomadism is a single form of food-producing economy which remains different from other foodextracting economies (for example, from different forms of hunting, fishing and gathering), and also from other food-producing economies (for example, from different forms of slash and burn and shifting horticulture, plough agriculture, etc.). However, it must also be admitted that pastoral nomadism is not completely separate economically from other foodproducing economies, rather it is linked with them by a series of transitional forms at the basis of which lies the gradually diminishing specific position of pastoralism. The appropriate boundary-line is not always drawn between these forms (see, for example, Kroeber, 1947:323-4; Markov, 1976:209; Marx, 1977:344-5) and this sometimes leads to a distorted picture bf nomadism. Thus the terms 'pastoral' and 'pastoralism', although they are very widely used, in themselves are too imprecise and unspecific. The basic forms of pastoralism At present there is no generally accepted classification of the different forms of pastoralism. Moreover, certain anthropologists sometimes even doubt that there is any need for or advantage in having one (see, for example, N. Dyson-Hudson, 1972:8; R. Dyson-Hudson, 1972:31, 47), evidently follow­ ing in this instance the principle of Leach (1961:3) which holds all those who devise typologies and classifications to be thoroughgoing 'anthropological butterfly collectors' (cf., however, Barth, 1966:22). To uphold their view they usually point to the existence of individual and group variations within the framework of the economy of the given society. Such doubts, although often based on field research, are not sufficiently convincing, for classifications of the basic forms of pastoralism examine the economy of a given society in its entirety and quite naturally concentrate on the principal and determining tendencies in that economy. For example, in 17

Nomadism

as a distinct form of food-producing

economy

many societies based on a pastoral semi-nomadic or mixed economy and even in certain agrarian societies there are families and groups which have nothing at all to do with agriculture, but who devote all their energies to pastoralism. However, we should be wrong to infer from this that such societies may be called pastoral nomadic. In effect, what we are encountering here is professional specialization and division of labour within one and the same society, while in a genuinely pastoral nomadic society division of labour between different societies is presupposed. It is true that since the beginning of the twentieth century geographers, particularly from France, have done a fair amount of work on the classification of pastoral economy (see, for example, Bernard and Lacroix, 1906, non vidi; Despois, 1949:217-325; Sorre, 1950:642-53; Veyret, 1951; Capot-Rey, 1953:251-2), but then there is no obvious prejudice against classifications in their subject. 'Classifications are basic in our teaching of geography' (Clarke, 1959:95). However, many of the classifications put forward by these geographers have a regional character. In addition, they reflect the principles and concerns of their authors' own fields of study and thus are not always and not on every detail acceptable to the anthropologist. Classification for the sake of classification is, of course, nonsense. Of course, no one anthropological classification can accommodate the entire multiplicity of specific cases and every kind of local and temporal specificity and already, therefore, should not be too rigorous or categoric. However, when classification is not an end in itself and makes for a more profound understanding of the essence of the phenomenon being examined and of its genesis, then it is useful. Without classifications and typologies many generalizations are impossible, and without generalizations anthropological theories, general theories of nomadism amongst them, are also impossible. Different scholars have chosen different criteria for their classifications of pastoral economy, both on regional and on global levels: geographical distribution, composition of herds, distance of pastoral migrations, direction of pastoral migrations, periodicity of pastoral migrations, character of dwellings, degree of sedentarism, character of sedentarism, specific role of agriculture in the system of economy, etc. (along with works already cited see, for example, Arbos, 1923:559-61; Patai, 1951:401-14; Bacon, 1954:55; Barth, 1962:342-3; Nomadic Pastoralism as a Method of Land Use, 1962:358; Jacobs, 1965:145-9; Whyte, 1966; Poucha, 1968:1212; Rathjens, 1969,20-8; Johnson, 1969; Rudenko, 1969:15-26; Murdock and Wilson, 1972:256-7; Baxter, 1975:207; Khazanov, 1975:5-8; Khazanov, 1978b; Simakov, 1978; Andrianov, 1978:123). The historiographical survey alone of all this would take up at least one whole chapter of the book. In principle the majority of such classifications cannot be considered either true or false. The main question to ask is how adequately they serve the specific aims of specific research. Hoping to deflect one possible 18

The basic forms of

pastoralism

reproach,'. . . that the typology makers never explain why they choose one frame of reference rather than another' (Leach, 1961:3), I stipulate in advance that, in accordance with the theme of the present book, the gradually diminishing specific significance of pastoralism in the general balance of the food-producing economy of traditional societies forms the basis of the classification proposed. I attach greater importance to the ratio of pastoralism and agriculture in an economic system than, for example, to degree of mobility for the very reason that apart from mobile pastoralism, there are mobile forms of agriculture; these include forms of the slash and burn type, together with those based on the use of arable tools, fallow crop rotation, including the three-field system of agriculture. The latter kind of mobile agriculture, for example, existed amongst the Eastern Slavs right up to the nineteenth century (Shennikov, 1971). The list of specific forms of pastoral nomadism is almost endless. However, the following stand out as the basic forms. Pastoral nomadism proper which in its most pure manifestations is characterized by absence of agriculture, even in a supplementary capacity. Despite some still existing opinions to the contrary, pure nomads are only recorded in certain regions of pastoral nomadism (North Eurasia, High Inner Asia, the Eurasian steppes, Arabia, the Sahara), but even in these regions pastoral nomadism co-exists as a rule with other forms of pastoralism. Semi-nomadic pastoralism is much more widespread through­ out the world. Semi-nomadic pastoralism is characterized by extensive pastoralism and the periodic changing of pastures during the course of the entire, or the greater part of the year; but although pastoralism is the predominant activity, there is also agriculture in a secondary and supplementary capacity. However, in North Eurasia semi-nomadic pastoralism is associated with other kinds of economic activity (see pp. 41-2). Even limited occupation with agriculture exercises a considerable influence on many aspects of the life of semi-nomads, in particular on the species-composition of herds, the routes and seasonal prevalence of pastoral migrations, etc. (see, for example, Khazanov, 1975:llff. on the seminomads of the Eurasian steppes; Doughty, 1888,1:234; Pershits, 1961:37-9 on the semi-nomads of Arabia; Evans-Pritchard, 1949:35; Clarke, 1959; Johnson, 1969:105, 109, 112-13, 115, 119, 130,142, 148,164, 168, 174, 175 on the semi-nomads of North Africa and the Sahara; Cunnison, 1966:22 on the Humr Baggara; Barth, 1964b:15-22; Swidler, 1973:33-7; Spooner, 1975:176 on the semi-nomads of Baluchistan; Ferdinand, 1969:143; 2

' It is curious to note tlmt some scholars who express doubts about the applicability of such a criterion nevertheless use the criterion quite widely in practice, although not always consistently (sec. lot cxnmplo. Inlmsim, 1969:17, 94, 97, 106, 148, 164, 170, 174, 175).

19

Nomadism

as a distinct form of food-producing

economy

Ferdinand, 1969a: 108ff. on the semi-nomads of Afghanistan; Kussmaul, 1962:223 on the semi-nomads of Tibet). There are many different variants of semi-nomadic pastoralism, as indeed there are of all forms of pastoralism, in the different regions of the oikoumene. It would appear that the two most important variants are as follows: (1) when the same groups in a given society (or sub-society) are occupied with both agriculture and pastoralism; (2) when within the framework of a given society (sub-society) there are groups which devote themselves primarily, or even exclusively, to pastoralism, alongside groups which are primarily occupied with agriculture. In almost all the regions inhabited by semi-nomads both variants are to be found: (see, for example, Orazov, 1975:215,216 on the Turkmen of Akhal; Capot-Rey, 1953:220 on the semi-nomads of the Sahara; Evans-Pritchard, 1949:35 on the semi-nomads of Cyrenaica; Fazel, 1973:131 on the Boyr Ahmad in Southwest Iran; Ferdinand, 1969:147 on the semi-nomads of Afghanistan; Barth, 1956:1082 on the Kohistanis). In the Libyan (Western) desert of Egypt the Jawabis tribe as a whole participates in both agriculture and pastoralism, and also in the transporta­ tion of goods. But in the same desert there are other tribes which are divided into two sections: one section leading a relatively sedentary way of life, the other a fully nomadic one. Nevertheless, both sections are parts of a single unity under one sheikh (Awad, 1962:334). Fukyi (1970:108-9,121), having studied the Iraqw of North Tanzania, singles out 'agrico-pastoral economy' as a distinct form in which, in contrast to a mixed farming economy, both elements of economic activity are separated from one another. Evidently, however, this is a concrete manifestation of the second variant of semi-pastoral nomadism. An analogous picture in East Africa may be observed, for example, amongst the Pokot with the one difference that the latter display both variants of semi-nomadic pastoralism (Schneider, 1957: 279; Schneider, 1964:70; Porter, 1965: 409-20; Conant, 1966:505-7). One variety of the first variant of semi-nomadic pastoralism is the situation in which the men move with the livestock and the women remain in one place, occupied with agriculture. Again in East Africa this is recorded, for example, amongst the Jie and the Karimojong, and is reflected in the proverb of the Jie, 'Grain is the stock of women' (Gulliver, 1955:61; R. and N. Dyson-Hudson, 1969:79). Again, amongst certain groups of the Somali, the men are occupied with pastoralism and the women with agriculture (Lewis, 1955:88); to some extent this state of affairs is also characteristic of the Fulani (Stenning, 1959:7), the Teda (Johnson, 1969:151-2) and certain other groups of pastoralists in the Sahara (Capot-Rey, 1953:220), etc. From an economic point of view fully nomadic and semi-nomadic pastoralism are especially closely connected and often interdependent, forming many transitional states which depend on local specificity, their 20

The basic forms of

pastoralism

specific historical situation and, of course, ecological conditions. For example, there exist between pastoral nomadism and semi-nomadic pastoralism, in which agriculture is a secondary but regular occupation, transitional forms; in these forms occupation with agriculture is irregular and not necessarily annual. It is recorded, for example, amongst the Beja, in Libya, Southern Algeria, Southern Morocco, Mauritania, the Sahara (Paul, 1954:13; Capot-Rey, 1953:220; Capot-Rey, 1962:302; Nicolaisen, 1963: 184). In some situations semi-nomadic pastoralism can be a relatively stable economic system and function for a long time in a more or less immutable form; in other situations it can be a transitional stage between pastoral nomadism and a mixed economy; and in others, alternatively, it can be a stage in development from a mixed economy to pastoral nomadism. Amongst many of the peoples of the Eurasian steppes who traditionally have been called, and still are called, nomads - for example, the Scythians, Hsiung-nu, Khazars, Tatars of the Golden Horde, and especially the Crimean Tatars, Turkmen and many others - there were semi-nomadic, sometimes very stable, groups (Rudenko, 1969:17; Khazanov, 1975: 11-12). Amongst the Fulani (Stenning, 1965:366) and the Baluch (Swidler, 1973:37), change in the semi-nomadic economy could take place in both directions, but in Tibet semi-nomadic pastoralism only developed into a purely nomadic form (Ekvall, 1968:22-3). However, even amongst semi-nomads agriculture is only a supplementary form of the economy, according to the definition of Krader (1959:505), 'a supplementary form of subsistence', which as a rule does not meet completely corresponding demands. Consequently, from an economic point of view semi-nomadic pastoralism is also dependent on the outside world. In certain regions, most particularly the Eurasian steppes and the Near and Middle East, it may be expedient additionally to separate semisedentary pastoralism from semi-nomadic pastoralism. Without becoming involved in the nuances of classification which are useless when an attempt is being made to propose a general classification, suffice it to say that semi-sedentary pastoralism more or less corresponds to what Forde (1963:404) has called 'sedentary cultivators with the auxiliary herding', and even more to what is sometimes called 'mixed farming' (Barth, 1976:75). Semi-sedentary pastoralism differs most fundamentally from seminomadic pastoralism in that in the former agriculture plays the predominant role in the general economic balance. Semi-sedentary pastoralism also implies the presence of seasonal migrations and/or separate, primarily pastoral groups and families within the given society. However, these migrations often seem to be shorter in both time and distance than the pastoral migrations of semi-nomads in the same kind of environment. In 21

Nomadism

as a distinct form of food-producing

economy

addition, semi-sedentary pastoralists are often involved to a limited extent in the laying-in of fodder. Thus, for example, in the summer the majority of the Karakalpaks living in auls used to move to pastures where they would live in yurtas; often these pastures were situated very near the aul, as little as a few hundred metres away {Narody Srednei Azii i Kazakhstana, 1962:440-1). Amongst the semi-sedentary Kirghiz at the end of the nineteenth and the beginning of the twentieth century, migrations only took place over three or three and a half months; moreover, not all members of the family joined in with these migrations, some remained sedentary all year round, occupied with agriculture (Simakov, 1978:24; cf. Ferdinand, 1969:129 on the semisedentary pastoralists of Afghanistan). Herdsman husbandry or distant-pastures husbandry describes the situa­ tion in which the majority of the population leads a sedentary life and is occupied for the most part with agriculture, while the livestock or, more often, some of it, is maintained all year round on pastures, sometimes quite far from the settlement, and tended by herdsmen especially assigned to this task. For part of the year the cattle are usually kept in enclosures, pens and stalls, which entails the laying-in of fodder. In the German tradition this form of pastoralism is sometimes called Teilnomadismus (see, for example, Poucha, 1968:121; Kussmaul, 1969:32). The specific character of distant-pastures husbandry is not only contained in the fact that entire groups of peoples no longer have to follow their livestock. Indeed, as it turns out, amongst some nomads and semi-nomads the livestock and the majority of the people move by different routes (see, for example, Gulliver, 1955:125-6 on the Turkana or Asad, 1970:19 on the Kababish). What is most important is that this form of pastoralism enables the majority of people who live by it to lead a continuously sedentary way of life. In societies in which herdsman husbandry is present, agriculture usually is more or less predominant; but in these societies pastoralism is still an important branch of the economy, sometimes so important that the economy is almost a mixed one. Since ancient times herdsman husbandry has been widespread in the Eurasian steppes, Arabia, North Africa and the Middle East. In the not so distant past many groups of Uzbeks drove their sheep out to pasture in spring and summer, and kept them in pens near the kishlaks in autumn and winter. Herding skills were handed down from generation to generation in specific families {Narody Srednei Azii i Kazakhstana, 1962:221-2). Hassan ibn Thabit, panegyrist of Mohammad, by birth a member of a sedentary tribe in the Medina area, was one of the first to ignore the traditions of aancientArab poetry and, instead of describing the Bedouin way of life, he (old of his own tribe in which there was pastoralism clearly akin to herdsman husbandry. i

22

,

The basic forms

ofpastoralism

Our tribe occupies a stony place surrounded by mountains in which distinguished people have built their dwellings and made a life. Palm trees grow here and between them canals have been dug out of the sandy, stony soil and there is a protective rampart . . . If you come to this stony place you will find a stall in which there are animals for riding and a herd of freely grazing domestic animals, (quoted in Filshtinsky, 1977:147).

Ibn Bishr, a chronicler of nineteenth century Najd, has left a description of herdsman husbandry in the oases of Arabia at the time of the first Wahhabi state. In spring the inhabitants of Sudair leave their camels, thoroughbred and [ordinary] horses [out in the pastures], they hire a man, he waters the animals, visits their masters, returns to the animals who are in their same place. He sees to their hobbles and tethers, then he leaves them, (quoted in Vasilev, 1967:200-1).

In Iranian villages there are often special herding groups, the members of which tend the livestock in desert and steppe regions over three or four months in the year (Barth, 1962:342; see also de Planhol, 1966a:295-6, 303-4 on the Lurs and Iranian Azerbaijanians; Spooner, 1976 on the Sangsari). The specific significance of pastoralism is usually at its most apparent in the specialized mountain variant of herdsman husbandry; in Soviet anthropology this is often referred to as yaylag pastoralism (from the Turkic word yaylag, meaning summer highland pasture). Yaylag pastoralism enables people occupied with agriculture in specific ecological zones to use other areas as seasonal pastures when they are at their most productive (Barth, 1962:342). During one part of the year the livestock is kept in mountain pastures and during the other parts is driven to lower zones. In Western anthropology yaylag pastoralism more or less corresponds to the notion of transhumance {transhumance, Transhumanz); see, for example, the definition suggested by Krader (1966:409). However, as it has been pointed out by Johnson (1969:18-19), many scholars often confuse transhumance with vertical variants of pastoral nomadism and seminomadic pastoralism; in doing this they ignore, on the one hand, the etymology of the term which comes from Spanish and was first used to describe specific forms of pastoralism in the Pyrenees, Alps and other mountainous regions of Europe (Sorre, 1950:647) and, on the other hand, the more essential fact that the vertical movement of livestock in itself signifies not a form of pastoral economy, but only some of its separate characteristics (Bacon, 1954:54). Sometimes the term is used even more broadly and less justifiably to describe all kinds of seasonal pastoral migrations in different forms of pastoralism, particularly those migrations which are not purely nomadic, or 23

Nomadism

as a distinct form of food-producing

economy

for describing seasonal utilization of different ecological niches (see, for example, Krader, 1955:302; Krader, 1959:500; Gulliver, 1955:18, 29; Bremand and Pagot, 1962:320, 324; Barth, 1962:342; Bohannan, 1966:48; Lattimore, 1967:309n.56; Klima, 1970:18). There are different variants of yaylag pastoralism, some of which are similar to semi-nomadic pastoralism, although most are similar to herdsman husbandry (see, for example, data on the relatively thoroughly studied pastoralism of the mountainous areas of Europe and the Caucasus: Arbos, 1923:562-4; Haufler, 1955; Viehzucht und Hirtenleben in Ostmitteleuropa, 1961; Krandzalov, 1963; Matley, 1968:231-61; Gamkrelidze, 1968:237-47; Viehwirtschaft und Hirtenkultur, 1969; Mandybura, 1972:7-9; Ismailov, 1973:59; Mambetov, 1974:134ff.; Klepikova, 1974:5-36; Robakidze, 1978:15-18; Shamiladze, 1979). However, in the Eurasian steppes, the Middle East and North Africa yaylag pastoralism often co-exists with semi-nomadic pastoralism and pastoral nomadism (see, for example, Janata, 1962/63:95-6; Schurmann, 1962:68-70, 277-8 on the northern Taimannis of Afghanistan; Coon, 1978:215 on the Kurds; Gellner, 1969:31-2, 207-8 on the Berbers of the High Atlas). Sedentary animal husbandry in primitive and traditional societies is not, as a rule, the basic form of economy. It only supplements agriculture and, in common with other forms of pastoralism and stock-breeding, has different variants. One of these variants is household-stable animal husbandry. Depending on climatic conditions, time of year, specialization of the economy, its technological level and a series of other factors, this form of stock-breeding is characterized by a situation in which for part of the year the livestock graze in pastures adjacent to the settlement (the animals usually return to the place of settlement every day), and for the other part of the year the livestock are kept in stables and enclosures and fed accordingly. If this variant of pastoralism is further subdivided into its many different varieties, the specific role of stockpiled fodder in maintaining livestock can serve as the main criterion for the subdivision. On one level, therefore, are primarily extensive varieties and, on the other level, there are semiintensive, or even predominantly intensive varieties. Yet another variant of sedentary animal husbandry is sedentary household husbandry, or stock-breeding, with free grazing. This is one of the most primitive forms of pastoralism in which the laying-in of fodder and the maintenance of livestock in enclosures or stables is generally absent, or if it is present it is only to a very limited extent. Although sedentary household husbandry with free grazing was geneti­ cally almost the most ancient form of pastoralism, it was able In exist in agricultural societies, much later, along with more developed forms. In 24

The species-composition

of herds

Russia, for example, it even lasted into the fourteenth, fifteenth and sixteenth centuries (Shapiro, 1977:72). Characteristic of this form was the small number of livestock and the minimal amount of time given over to taking care of them. In conclusion I must point out that the suggested classification is not only the fruit of theoretical work for, to a certain extent, it is based on traditional ideas and criteria of pastoralists themselves over a number of centuries. In a document from Bukhara of the end of the sixteenth century reference is made to the difference between inhabitants of villages, kishlaks and the steppes, i.e. between sedentary agriculturalists, semi-nomads and nomads (Barthold, 1963d:467). The Arabs differentiated between purely nomadic camel-breeders, sheep-breeders who were sometimes really semi-nomads, semi-sedentary pastoralists and agriculturalists (see Musil, 1928:44-5; Dickson, 1951:108-11; Coon, 1976:198-9). The forms of pastoralism which have been examined are also different methods of economic adaptation which, in the final analysis, have been determined by the sum total of very diverse factors. At the same time it is important to bear in mind that in specifically functioning societies these forms are not absolutely static; on the contrary, they merely point to the parameters of a changeable economy which is capable of transformation, they can succeed each other on the same territory and this does not always happen in identical sequence. However, in the opposition between nomadism and sedentarism, according to which the essence of many of the forms of pastoral economy and its changes is often defined, the agriculture at stake is one of the most important criteria. In many of their different characteristics and features pastoral nomadism and, even more so, other forms of extensive pastoralism demonstrate a remarkable diversity. In the present work there is no room or necessity to examine all of their features and characteristics. What interests me here is how they can be used as criteria for a typology of pastoral nomadism based on its geographical distribution. The species-composition of herds The species-composition of herds, their structure, i.e. the sex- and age-ratios of the animals and the way in which the herding-animals are employed, are determined, first and foremost, by their biological particular­ ities and natural geographical conditions. But also influential are economic, social, political and cultural factors, at best only loosely connected with the ecology (Jones and Whittlesey, 1932:379; Forde, 1963:400). In Mongolia, in the fifties of the present century, in an identical ecological situation, well-to-do households kept more large stock, while poorer households were involved with small stock; the well-to-do were more 25

Nomadism

as a distinct form of food-producing

economy

involved with sheep-rearing, the poorer with goat-rearing (Zhagvaral, 1974:98). Many Mongols struggled to own more horses than were strictly needed in their economic circumstances. In 1948 one informant offered an explanation for this. 'Since I have horses I give them to others to use on journeys, to help themselves to koumiss and thanks to this people feel kindly towards me and give me whatever help I need, particularly in pastoral migrations, looking after stock, sheep-shearing and taking the wool to the co-operative, in fulfilling urton service [i.e. postal and transporting duties A.K.]. This is the merit of the horse' (Zhagvaral, 1974:99). The Bedouin of Arabia kept horses because they were prestigious and could be used for fighting, although horses were clearly unsuited to the environment there; in fact, the Bedouin were almost as proud of the pedigrees of their horses as they were of their own genealogies. According to Stauffer (1965:292) the proportion of sheep and goats kept by nomads in Iran depends on how far away there are markets. The closer the markets the more sheep there are in the herds, but the further away the markets, the more goats in the herds, for there is less demand for goats in the markets and, at the same time, goats are less trouble to tend and give 50-100% more milk. All hoofed animals (ungulata), including the wild forefathers of nomadic herding-animals, were able to disperse quickly over the oikoumene at the end of the Ice Age, or again when trees disappeared because of climatic changes. The distinguishing feature of these ungulata, which has consoli­ dated and even strengthened amongst their domestic descendants, is their marked ability to adapt themselves to different living conditions and to radically alter their behaviour, even in the first generation, if their environment changes (Baskin, 1976:187). Nevertheless, the distribution of ungulates is still subject to ecological constraints. Thus, reindeer can only be herded successfully in the tundra, camels in the desert and yaks in high mountainous regions. For this reason the Arab tribes of Southern Kordofan and Darfur had to abandon camel herding as conditions in the savanna were unfavourable (Awad, 1962:330). Another example relating to the camel is apposite here. The idea that camels are undemanding about what food they eat in actual fact only applies to desert plants which are rich in salt. When camels raised in desert conditions feed in rich pastures they grow thin and finally die (Baskin, 1976:71). For exactly the same ecological reason, lack of food, attempts to introduce zebu into the herds of the Tuareg of Ahaggar failed (Nicolaisen, 1963:49). Of course, the adaptability of such animals as sheep, goats, partly even of horses and cattle is far greater than that of relatively specialized species of 3

3

koumiss - fermented mare's-milk.

26

The species-composition

of herds

animals; for this reason the geographical distribution of the first group is significantly wider. However, the adaptive capacities of these animals are one thing, the economic expediency and effectiveness of herding them in specific ecological conditions are another. Both goats and sheep are able to pasture in the same ecological zones, and in the Eurasian steppes, in the Middle East and in North and East Africa they often pasture together. In this situation a billy-goat plays the role of leader. Goats are less demanding about the food they eat, more easily adapt to changes in temperature and are more hardy; however, they do not like humidity (Dahl and Hjort, 1976:250-1) and their produce, the wool of certain specialized breeds excepted, is of poorer quality. The Tuareg readily acknowledge the biological differences between these species and in certain situations, for example, in Southern Ayr, they herd more sheep than goats, while in other situations, for example, in Tassili, goats predominate among small stock (Nicolaisen, 1963:45-6). With regard to species-composition of herds, pastoral nomadism can be subdivided into the monospecialized and multispecialized. Whilst the nomads of the extreme North herd only reindeer, most other nomads in the Old World (on the problem of nomadism in the New World see p. 68) herd different species of animals at the same time although, of course, in different proportions. The fact that monospecialization itself can take place on different levels should not just be ignored in the way that it is ignored by Krader (1959:500). Amongst reindeer-herders monospecialization is char­ acteristic of the nomadic society in its entirety. In the Near East it is sometimes inherent in separate nomadic groups which are integrated into a larger system or supersystem. Uniquely specialized types and variants of pastoral nomadism are almost always imposed by ecological necessity. Multispecialized types relatively are more stable and lasting, and permit wider utilization of pasture. The loss or reduction in numbers of one species of livestock can be compensated, to some extent, when other species are available, because capacity to resist different natural disasters, and reproductive abilities, vary amongst different species of animals. On the other hand, alternation between periods of relative prosperity and adversity, which are linked to natural and biological factors, are especially characteristic of monospecialized pastoral nomadism. The biological particularities of herding animals which nomads have to take into account have a bearing not only on the species composition of their words, but also on ways in which the herds can be split up and pastured in specific geographical conditions. In the extremely arid con­ ditions in which the Bedouin of Arabia, the Tuareg of the Sahara and Hie Somali of the Horn of Africa live, camel-rearing, on the one hand, and sheep- and goat-rearing, on the other, are rarely compatible and usually differentiated in different areas of economic activity. However, the 27

Nomadism

as a distinct form of food-producing

economy

nomads of the Eurasian steppes often have their horses, sheep and large stock grazing together (see p. 49). The composition of nomads' herds is naturally determined first and foremost by the biological particularities of the animals being pastured. Nomads can and, indeed, often do resort to essential corrective action in regard to the composition of their herds, and this differentiates their herds from herds of wild ungulata. A herd of domestic animals is, for example, much younger than a herd of its wild counterparts. The general lack of old animals in the conditions of pastoral herding leads to a lack of potential leaders and, consequently, man must become involved directly in leading the herd (Baskin, 1976:271). The number of productive and unproductive animals in a herd can be regulated according to the demands of the given group of nomads for the various products of pastoralism and different kinds of the latter, also for activities which are not strictly pastoral, transport, for example (see Cunnison 1966:38 on the Humr Baggara), and for various cultural reasons (Spooner, 1973:12; Digard, 1973:45-8; Dahl and Hjort, 1976:281, 32). Finally, different groups of nomads use both the products of pastoralism and their animals in different ways. Amongst the Bedouin of Arabia camel-herding is an important source of food; for the nomads of the Eurasian steppes and the Middle East camels are used primarily for transport, most particularly as beasts of burden; and the Turkana regard camels as a type of cow and never use them for transport (Gulliver, 1955:39, 260-1). For the majority of the Tuareg of Ahaggar the products of camel-herding are a relatively unimportant source of food; but in the past camels were indispensable riding animals, used in raids as well as in caravan trading, from which the Tuareg procured most of the vegetable foods they needed in their diet (Nicolaisen, 1963:404; Lovejoy and Baier, 1976:155). However, there exists no clear and simple dependence between the different forms of extensive pastoralism and the species-composition of herds. All that can be said is that within the boundaries of one region the specific forms of pastoralism are usually connected with, or conditioned by the specific species there. Thus, as a general rule, in the desert and steppe regions of Eurasia camels and horses are most characteristic of those forms of pastoralism which are less involved with agriculture, and large stock (cattle) is characteristic of those forms more involved with agriculture. But in other conditions, in East Africa, it is large stock (cattle) which plays the leading role in the herds of the nomads. Size of herds In recent years anthropologists have focused a great deal of attention on this problem and on a whole range of related problems. However, the practical 2H

Size of herds usefulness and applicability of such basic notions as the maximum and minimum size of herds should not be exaggerated (Khazanov, 1980). According to Spooner (1973:9) and some other anthropologists,'. . .the ability of the herdsmen to control the animals in any given topographical situation is a factor in determining the maximum size of the herd. The requirements of the family or other grouping that subsists from the herd are a factor determining the herd's minimum size.' First of all it must be pointed out that the two notions, although they would appear to stand in interdependent opposition to one another, relate to different spheres of pastoral economics and thus are not really inseparable. 'The maximum size of the herd' relates to the technique of effective management of the herd, and 'the minimum size of the herd' to the necessary subsistence minimum. The number of animals which practically can graze together usually depends on a complex combination of ecological factors connected with environment, weather conditions in the given year and even in every season of the given year, epizootic circumstances, species-composition of herds, biological particularities, and the age- and sex-structure of herds, etc., and also on socioeconomic factors such as the number of livestock at the disposal of each separate household, size of the available work force, type of grazing and utilization of animals, etc., and, finally, even on cultural traditions and professional skills. Since the majority of those factors are changeable, both within the whole of a nomadic society and even within separate families and households, the maximum size of herds is no stable criterion; even less can it be used as a criterion for comparing different nomadic societies. Equally unhelpful is the notion of the minimum size of herds. Any specialist will acknowledge that any definition of a subsistence wage in a contemporary society is imprecise, vague and variable, and depends on the level of development in the specific society, on economic conjunctures, political considerations and even on methods of calculation. Suffice it to recall Disraeli's view that statistics are a form of lying. Of course, nomadic societies are not based on a money economy, at any rate they were not based on such in the past. However, the fact that nomadic economies are not autarkic means that the minimum size of herds partly depends on supplementary sources of subsistence; these sources vary from one nomadic society to the next, and even amongst the different families within these societies; they can also change from year to year depending on specific historical and economic circumstances. In this respect the observation of Ekvall (1968:19) that Tibetan nomads find it difficult to calculate the exact number of resources they need to subsist on is illuminating. It is not surprising then that the specific figures with which different scholars define the minimum size of herds amongst nomads vary very considerably, not only from region to region, but also within these regions. 2

Nomadism

as a distinct form of food-producing

economy

However, apart from subjective reasons connected with the inadequacy of calculations, the incompleteness or inauthenticity of information collected (even in traditional societies statistics can be a form of conscious lying), this state of affairs is also linked to the objective circumstances mentioned above. Here are some examples. The Eurasian steppes. According to Pallas (1776:226) a Kalmuck family of 5 in the middle of the eighteenth century needed 8 mares, 1 stallion, 10 cows and 1 bull to survive. Zhitetsky (1892:212) thought that the minimum herd amongst the Kalmucks was 50-100 head of livestock for each family. Georgi (1776:125) recorded that a fairly well-off Kazakh family kept 30-50 horses, 100 sheep, 15-25 head of large stock, 20-50 goats and several camels. Slovtsov (1881:23-5) wrote that amongst the Kazakhs of the Akmolinsk district a family of 5 or 6 needed 5 horses, 10 rams and 6 cows to survive. According to different statistics a family of 5 could maintain their own household on 15 horses, 2 camels, 6 head of large stock and 50 sheep (Narody Srednei Azii, 1963:330). However, according to Tolybekov (1959:131) a Kazakh family of 4 to 6 people needed 15-20 camels with their young, 4-5 horses and 100-150 head of sheep and goats. Below this minimum even simple reproduction could not take place. Maisky (1959:140-1) thought that a Mongol family of 5 at the beginning of the twentieth century needed 14 horses, 3 camels, 13 head of large stock and 90 sheep and goats. According to the calculations of Rudenko (1969:18), to satisfy its minimal needs a family of 5 needed the number of livestock which was the economic equivalent of 25 horses; he calculated that 1 mature horse equals 6/5 head of large stock, and 6 sheep and goats. Furthermore, other animals were needed to transport dwellings and goods, and a riding animal was needed for each adult member of the family. The Middle East. According to Pastner (1971:177) a family of nomads in Makran can lead a relatively well-to-do life by local standards if it has 1 camel or donkey and 30 sheep or goats, and exchanges some pastoral products for agricultural ones. Swidler (1973:40 n.13) has noted that the Brahui receive most of their monetary income from selling livestock and that elementary Brahui families possessing less than 30 mature sheep or goats are not uncommon. According to Barth (1964:16-17; 1964a:73) in South Iran an elementary family needs at least 60 mature sheep and goats to subsist with. Digard (1973:37) notes that amongst the Bakhtiari the subsistence minimum consists of 50 head of small stock. In Gharajistan the minimum ratio of livestock to one person is 10 sheep and 1 camel (Glatzer, 1977:38). Africa. According to Silberman (1959:569) amongst die Somali a woman with 3 children can survive on 50 sheep or goats and 1 milch camel, while 100 30

Size of herds head guarantee a reasonable subsistence level; another 2 animals are needed for transport. Lewis (1961:58) gives similar figures. According to Asad (1970:52) amongst the Kababish a nuclear family needs 20-25 camels or 40-50 sheep to subsist independently. Amongst the Karimojong a moderately well-to-do man has 100-150 head of large stock, 100 sheep and goats and a few donkeys (R. and N. Dyson-Hudson, 1970:107). Gulliver (1955:39) writes that amongst the Turkana a family of average means has 25-30 head of large stock and 100-150 head of small stock. If these figures are compared with the calculations of Dahl and Hjort (1976:266) according to which a 'referential family' of nomads should possess 50-64 head of large stock (cattle), or 28 camels, or more than 100 head of small stock (this does not take into account vegetable foods), or if they are compared with other calculations according to which the minimum subsistence level in cases where no cereal are grown must equal 2-4 standard stock units (SSU) per person, it becomes quite clear if not that the reports of specific observers are unreliable then, at least, that they are nonrepresentative beyond the boundaries of the given group of nomads at the moment of observation. In the lives of nomads there exist so many changing factors, ranging from fluctuations in prices for livestock and vegetable produce to size of non-pastoral sources of income, that the notion of the minimum herding unit is of little use in many generalizations. Many scholars, pointing to the inadequacy and imprecision of the notions being examined, outline yet another one, 'the optimal size of the herd' which oscillates 'between the maxima and minima' (see, for example, Leeds, 1965:90). Of course, when deciding the optimal size of the herd nomads, with the practical experience of many centuries, take into account the biological and behavioural particularities of the animals they are pasturing. For example, the deliberate and constant maintenance of animals in large herds is extremely important because it permits the utilization of mechanisms of behaviour which only appear in dense masses of animals when they are rigidly controlled (Baskin, 1976:271). The well known herd instinct of sheep - remember Pan and his flock - is, evidently, the sheep's adaptation to conditions artificially created by man. However, neither is there any precise data on the optimal size of herds; furthermore, the optimal size would appear to be different in different regions. According to the calculations of Dahl and Hjort (1976:270) the optimal size of a herd of camels is 150; one herdsman without a dog can look after 400 head of small stock, and with the help of a dog he can look after up to 500 head. 4

4

From the calculation that a S S U equals 10 sheep or goats, 2 cows, 1 camel (Le Hou6rou, 1977: 25, 38 rt.2).

,11

Nomadism

as a distinct form of food-producing

economy

However, amongst the Rwala camels were pastured in herds of no more than 70-80 head (Musil, 1928:336). In Southwest Asia one shepherd can pasture a herd of 300-400 sheep even where the terrain is quite difficult (Barth, 1962:345; Trubetskoi, 1966:104). Amongst the Brahui the optimal herd is thought to be 500 head, and if a herd numbers less than 250 head then management conditions are thought to be bad (Swidler, 1973:28; see also Barth, 1964:22). Nevertheless, amongst the Bakhtiari a herd of small stock does not usually exceed 200 (Digard, 1973:45). Amongst the Kalmucks in the nineteenth century two herdsmen were able to look after a flock of sheep of 1,000-1,500 head or a herd of 300 horses (Zhitetsky, 1892:95-6). Amongst the Turkmen of the not very distant past a herd of 400-800 head of small stock was tended by one shepherd and a herdsboy (Orazov, 1970:202); a flock of up to 800 head of sheep was thought to be manageable, and more than 1,000 head to be inconvenient and difficult to pasture (Babadzhanov, 1975:229n.l6). In Inner Mongolia one person on foot could cope with a flock of 150-200, or even 500 sheep if he had a horse; two horsemen could cope with 2,000 sheep; and one person could pasture a herd of 150 horses. 'But you seldom see a flock of more than 1,000 sheep, because sheep have a habit of eating up all the grass without leaving anything and so those in the rear can hardly find any grass to eat' (Goto, 1970:95-6). Amongst the reindeer-herders of Kamchatka and Chukotka the maximum size of the herds is 1,000-1,500 head; the herders here are able to cope with larger herds, but pasture deteriorates when the number of animals grazing on it exceeds this maximum. If a herd is too large it can no longer be accommodated in the narrow valleys of the tundra, nor can it utilize the small areas on the tops of hills and windswept spots where the snow is less deep and hard packed (Baskin, 1976:81). Economic, social and sometimes even ecological needs and necessities often lead nomads into making specific and practical choices, rather than ideal ones, when they decide the optimal size of their herds. Cunnison (1966:63, 67n.7) cites a good example of this phenomenon. Although the Humr Baggara have some conception of the optimal size of their herds, in practice the size of the herd often depends purely on what work force is available. In effect, the optimal size of the herd per se in many ways depends on the technique of pasturing and certain other non-biological factors. Contempor­ ary ranch farming serves as a good example here. In the USA one horseman can cope with 1,000 head of large stock. In Australia up to the fifties of the nineteenth century one person was needed to tend between 300 and 1,500 head of sheep, afterwards one person could cope with as many as 2,500 head. In Argentina 2½—3 people usually are needed to cope with 1,000 head of large stock, and 1½ usually needed to cope wilh 1,000 head of sheep and 32

Nature of the utilization

of ecological

zones

goats (Strickon, 1965:245). At the same time, according to my field data, in 1972 in the sovkhoz (state-owned farm) 'Zapadny', one of the most efficient in Kalmuckia, where sheep are maintained all year round on a system of free-range grazing and water is taken out to the dry pastures in tanks, a flock of 850 is tended by 4 shepherds. In no way am I trying to say that it is unnecessary to study the economic, sex and age, size or any other factors about the herds of nomads; however, what I am saying is that our present knowledge of these factors is still very incomplete and that the employment of whatever knowledge we have in a comparative study of nomadic societies, particularly in a study of the interrelations between nomads and the outside world, is no easy task. Nature of the utilization of ecological zones The variety of specific cases in this section may be divided somewhat systematically into three basic variants. (1) Nomads occupy the most exclusive or the dominating position in the zone they inhabit and, to whatever extent this is possible, sometimes try to utilize neighbouring zones, primarily marginal areas. In this respect the Eurasian steppes, semi-deserts and deserts are a good example, for here there is a relatively clear-cut geographical and ecological demarcation between zones of pastoral nomadism, and oases, i.e. zones where there is irrigation agriculture. Although de Planhol (1970:445) maintains that in High Central Asia agriculture has been possible wherever there has been pastoral nomadism, this is in fact not the case at all; oscillations in the use of one or other form of economic activity have only occurred in marginal areas (see pp. 44-5). Another example showing how nomads occupy the dominating position in the ecological zones they inhabit is provided by the reindeer herders of the Far North. They very quickly occupied vast areas of the tundra, and right up to the twentieth century continued to extend the territory in which they herded their animals (see pp. 72-3, 80). In this respect the nomads of High Inner Asia are not very different; in effect, the one difference is that here the geographical demarcation between nomads and agriculturalists is altitudinal rather than latitudinal. Within this first variant may also be included those areas of Arabia and, more particularly, of the Sahara where there is agriculture only by oases, which constitute a small incursion into territory largely occupied by nomads. (2) Nomads utilize several ecological zones which are separated by other zones, in which there are people engaged in different economic activities, but usually in agriculture. Nomads themselves do not utilize these zones, but have to travel across them during pastoral migrations. This can create certain complications for both sides and, in fact, the situation is often further 33

Nomadism

as a distinct form of food-producing

economy

complicated when ecological zones are utilized on a rotational basis either by different pastoral groups, or even by groups in which there is a mixed economy. This second variant is most widespread in the Middle East, and to a lesser degree in the Near East (Barth, 1959-60; Barth, 1962:342; Ferdinand, 1962:124, 129-30; Johnson, 1969; Swidler, 1973:25; Marx, 1977: 351, 358). It is not, however, uncommon in the Eurasian steppes. The sandy desert near Krasnovodsk was utilized by the Turkmen in summer, and the Kazakhs in winter (Ishchenko et al, 1928:81). (3) Nomads do share, either fully or partially (much more often the latter), the same zones with agriculturalists. There are different variations of this situation. Sometimes one part of the territory is utilized exclusively by the nomads, another by the agriculturalists, and another by them both, usually on a rotational seasonal basis. Sometimes nomads monopolize their own ecological zone and partially utilize a zone of agriculturalists into the bargain. Sometimes, on the other hand, all the nomads (or pastoralists) occupy is their own separate niche in the ecological zone of agriculturalists. In all such variants co-existence is usually characterized by nomads pasturing their livestock on fields which have been harvested. Sometimes nomads pay agriculturalists for the use of the stubble, sometimes they themselves receive payment for the manure they are providing. In some regions such a relationship between pastoralists and agriculturalists is more ancient than pastoral nomadism; it is recorded, for example, in the second millennium B.C. in Mesopotamia in the reign of Mari. As a whole this variant is most characteristic of the nomads of the Middle East (see, for example, Bates and Lees, 1977:833 on the Turkish Yoriik; Coon, 1976:216 on the Kurds; Narody Serednei Azii, on the Gilakis of Gilan; Trubetskoi, 1966:11 on the Bakhtiari; Schurmann, 1962:262 on the nomads of Afghanistan; Barth, 1964b; Swidler, 1973:24 on the nomads of Baluchistan). However, it is fairly widespread in the Near East (see, for example, Pershits, 1961:31 on the Bedouin of Arabia; Fernea, 1970:12 on the camel-herders of Southern Iraq; Johnson, 1969:105 on the Said Atba in Algeria) and in the Sudano-Sahelian zone (Stenning, 1959:6; Thomson, 1977:60; Gallais, 1977; Frantz, 1978:102; Dahl and Hjort, 1979:7). July (1975:160) assumes that it was the Hausa who invited Eulani pastoralists into their territory in order that their fields should be fertilized with manure, which the herds could provide; at the same time this assisted the pastoral migrations of the Fulani. In the Eurasian steppes this variant was less common, and mostly could be found in places where there was seminomadism and herdsman husbandry (see, for example, Sazonova, 1978:68). Different examples of the ways in which nomads share ecological zones with agriculturalists are best and most comprehensively d e s c r i b e d by Barth (1956; 1959-60; 1964; 1964b). However, despite individual reservations, in 34

Nature of the utilization

of ecological

zones

this respect his work suffers from a certain biologism. Nomads and agriculturalists essentially become like different species of animals which utilize natural resources in different ways; consequently, they do not compete with one another and sometimes even the relations between them are symbiotic. Political relations and property rights which, in the end, define the character of the joint utilization of a zone are insufficiently considered. Firstly, however, neither pastoral nomadism nor any other economic system are absolutely static, they are all capable of expansion even if this is at the expense of one another (see Barth's own data on Northwest Baluchistan, Barth, 1964b:18-19). Secondly, symbiosis entails not only co-existence, interaction and even interdependence, but also mutual benefit; however, relations between nomads and agriculturalists in many cases which have been examined do not correspond to this definition. Much more often they are a modus vivendi which is guided not only by mutual interest of a purely economic kind, but also by the correlation of forces both amongst directly involved groups and in the wider outside world where, in the end, the balance of these forces is determined. In interrelations such as these mutual benefit has rarely been the dominating and balancing factor (see, for example, the interrelations between the Hausa and the Fulani in the Sahel of Niger, Diarra, 1975). It is very difficult, for example, to follow Barth (1956:1083) when he describes the relations between the Pathans and the Gyjars in Swat as symbiotic if, as he writes, the former were landlords and the latter their clients, or serfs, or servants. But there are also less extreme cases. Between the pasturing of livestock on fields already harvested and the damaging of these same fields lies one short step; this step can be beneficial for nomads, but can be ruinous for agriculturalists. Medieval sources are full of testimonies of how the agricultural population of the Near and Middle East suffered from the nomads' practice of driving their livestock across fields sown with crops (see, for example, Reisner, 1954:43). Indeed, this was still common practice not so long ago in Iran (M. S. Ivanov, 1961:40). In the second half of the nineteenth century in Khuzistan the Lurs and Bakhtiari left no villages standing and no fields untrampled along their routes during the annual pastoral migrations (Lambton, 1953:157). On the other hand, the Yoruk had to pay an ever increasing rent (as agriculture developed in Turkey in the area adjacent to the coastline) for the right to pasture their livestock for a few months on the fields of the agriculturalists (Kolars, 1963:48-50; see also de Planhol, 1959:529). During their pastoral migrations from spring to summer pastures the Kurds of Southeast Anatolia were forced to drive their livestock along incommodious routes so as to avoid collisions with agriculturalists (Johnson, 1969:37). With the introduction in the Sahel of cash crops, such as cotton, which are .15

Nomadism

as a distinct form of food-producing

economy

harvested late nomads were no longer able to pasture their herds on the stubble in the dry season (Norton, 1976:260-1). In all these cases 'symbiosis' has in fact turned into a simple correlation of forces whereby the stronger are guided exclusively by their own interests and needs, without any consideration at all of the interests of the weaker. A curious example of how difficult it is for pastoralists and agriculturalists to co-exist in one ecological zone is cited by Gellner (1969:33f., 170f.). The Berbers of the High Atlas are forced to resort to saintly intermediaries, who live in the marginal areas of their territory, for help in solving this problem. Even in the nineteenth and the beginning of the twentieth centuries in eastern districts of the Emirate of Bukhara sedentary and semi-nomadic groups (nomadic ones as well in earlier times) co-existed and often shared the same ecological zone, competing for the utilization of separate niches and enclaves within it. But here there was no intergroup division of labour founded on mutual benefit, rather there was compulsory co-existence accompanied by conflict and the struggle to supplant competitors (Karmysheva, 1976:145, 164, 190, 263). R. Cohen (1978b: 155) is quite correct when he writes: 'The structuralfunctional assumption of a balanced symbiosis between nomads and farmers turns out to be false when examined in historical perspective.' I could also have expressed my thinking in the language of pure ecology, in which case the following categories of interaction between two species would be singled out. (A) Symbiosis ( + , + ) - each of the species accelerates the growth of the other species. (B) Commensalism (+,0) - one species has the advantage without harming or helping the other. (C) Predation ( + , - ) one species obstructs the growth of the other thereby accelerating its own growth. (D) Competition (—,—)- each of the species obstructs the growth of the other. (In my description of these categories I basically follow Smith, 1974, with certain additions and amplifications). According to this language, the interaction between nomads and agriculturalists co-existing in one ecological zone, which Barth joins together under the general term 'symbiosis', should emphasize the entire range of categories from + , + to - , - , despite the fact that symbiotic relations (if, indeed, they do exist anywhere) are very rare. However, I am no advocate either of equating human societies with biological species, or of examining them only as elements in natural ecosystems. Human societies and natural ecosystems make up supersystems founded on reciprocal connections. Most importantly, the dif­ ferent variants of the interaction between nomads and non-nomads are far more numerous and varied than the interaction between two biological species. Thus once again I should like to emphasize the point that it is far from true that economic interdependence and mutual interest even can always be defined as symbiosis. Political relations which are lacking in the 36

Character of pastoral

migrations

animal kingdom colour very specifically the interaction between nomads and non-nomads, including the interaction between those sharing the same ecological zone. As a whole the second and third variants of the utilization by nomads of ecological zones are closer to one another than they are to the first variant. However, all three variants often supplement rather than exclude each other. For example, the Basseri, who as a whole belong to the second variant nevertheless make use of opportunities to pasture their livestock in autumn and winter on the fields of agriculturalists (Barth, 1959-60:7). Character of pastoral migrations The pastoral migrations considered in this section are those which nomads regard as necessary for the grazing requirements of their livestock rather than just mobility, which anyway is the way of nomads and can be for dif­ ferent economic and non-economic reasons. This fact has been noted many times by many different scholas (see, for example, Salzman, 1971:132-93). However, it is need of suitable pasture for the livestock which more than anything else determines the character and need itself of pastoral migrations amongst nomads. The Lokai Uzbek say: 'The food of stock is on the ground, if stock feed from human hands they will never be sated.' As with all the other criteria of nomadism, the character of pastoral migrations is determined by many factors, including those which are examined above. Thus, migrations are fairly adaptable and versatile. However, in all circumstances the specific character of the environment is one of the most important determinants. The character of the pastoral migrations of different groups of nomads varies very considerably, even amongst nomads within one region. Regularity and stability vary; cycles (interseasonal and in-season pastoral migrations) vary; distance varies; directionality varies, it can be both vertical and horizontal, either linear (which may be divided into the meridional and the altitudinal) or nonlinear (which may be divided into the elliptical, radial-circular, etc.); economic priorities vary, whether for food and/or water and/or temperature, pollution of the locality, etc. 5

6

7

8

9

* Hy regularity of pastoral migrations is implied their temporal aspect and by stability their spatial aspect. Vertical pastoral migrations are connected with changes of pasture in an altitudinal direction. Pastoral migrations which may be called elliptical are those whose annual routes of migration, when traced out on a map, form an ellipse or a similar geometrical figure. " Pastoral migrations which may be called radial-circular are those which move around one more or less stable centre, which is usually connected with a source of water to which the herds periodically return. An interesting and as a whole acceptable classification of the routes of pastoral migrations has been put forward by Johnson (1969). However, it is my view that Johnson has exaggerated the significance of this criterion in the typology of nomadism as a whole. Moreover, it would seem 1

v

37

Nomadism

as a distinct form of food-producing

economy

While interseasonal changes of pasture are determined mainly by natural and climatic conditions which allow for specific pastures to be utilized only at a specific time of year, in-season pastoral migrations usually depend primarily on the size of the given herd, the food and water it demands, and a series of other reasons. Interseasonal changes of pasture in themselves already imply a certain regularity and cyclic recurrence of pastoral migrations; moreover, as a rule they are much more protracted than inseason migrations. Clear distinction between interseasonal and in-season pastoral migrations is most characteristic of nomads in northern and temperate zones; in the hot desert zone the distinction is much less clear. The stability of pastoral migrations which are linked directly with the pasturing of the animals is determined mainly by the availability of food and water in the same places and at the same time of year. It is hardly surprising, therefore, that there is less stability of movement amongst many groups of nomads in Arabia, the Sahara and East Africa than there is amongst nomads of other regions. It should be stressed that there are no nomads whose pastoral migrations may be referred to in terms of absolute stability or of absolute instability; even the most firmly established routes are not necessarily repeated every year, while from time to time the most unstable routes coincide. At the same time it is true to say that when we look at the nomadic routes of the large subdivisions of a nomadic group, these routes seem to be more stable than those of the other, smaller subdivisions in the same group. Nevertheless, the greater or lesser stability of pastoral migratory routes can be used as one criterion for comparing nomadism in different regions. Length of pastoral migrations is affected mainly by seasonal availability of fodder and water, and also by size of herd, its species-composition and many other factors. As a result distances vary between migrations of a few score kilometres amongst the cattle-herders of East Africa to those of a thousand kilometers or more amongst nomads in Arabia, the Middle East and the Eurasian steppes, and certain groups of reindeer-herders. However, in this respect, not infrequently there is as much variation within individual regions as there is between different regions. Equally diverse is directionality of pastoral migrations, including the reasons by which directionality is determined. All that can be said here is that while vertical migrations everywhere are linked first and foremost with geographical and environmental conditions, they only predominate in High Inner Asia and in the mountainous areas of other regions. It is as difficult to generalize about the economic priorities of pastoral migrations as it is to do so about their other characteristics. Taking two of that he is wrong to make so much of the contrast between vertical and horizontal nomadism for, as he himself points out, the same nomads are often involved for a part of the year in horizontal pastoral migrations, while for the other part they are involved in vertical ones.

38

The main types of pastoral

nomadism 10

the most important economic priorities - fodder and water, all that can be said is that the problem of food is greater for nomads in northern and temperate zones and the problem of water is greater for nomads in hot zones. Account must also be taken of the fact that in specific nomadic societies other priorities play an important role. The problem of salt in the Sahara is one example (Capot-Rey, 1953:37; Smith, 1978:86). Careful consideration is given in the pastoral migrations of the Fulani to the distribution of the various types of tsetse fly in different seasons (Stenning, 1965:371-2) and in the migrations of Tibetan nomads to epidemics and the pollution of different regions by insects (Ekvall, 1968:19). However, on a regional level such priorities as a whole may be considered as secondary. Nature of the utilization of the products of pastoral economies and dietary systems Here both the nature of the utilization of the products of pastoral economies as food and, more extensively, the dietary differences between nomads of different regions should be considered. In common with all the other criteria this one also depends on many different factors, amongst which may be listed ecological situation, species-composition of herd, particularities of relations with the outside world, cultural, including religious and ethnic traditions, and many others. Unfortunately, nowhere in the relevant literature have I been able to find a definite answer to the question of whether a diet of animal products alone, without any vegetable supplements, in principle can be healthy and balanced. Judging from the large amount of data obliquely touching this question, it would appear that the reverse is more likely, at least it is a more widespread tendency (cf. Spooner, 1973:6). All, or almost all, nomads include vegetable foods in their diet, although in different quantities and they procure these foods by different means; in this respect there are fairly marked differences between different groups of nomads. There are also considerable differences in the ways in which they use the three basic products of pastoral economy - milk and milk products, meat and blood. The assertion of Dahl and Hjort (1976:23) that '. . . the primary goal for nomads is to produce milk for subsistence use' is scarcely applicable to nomads in northern zones, and to some extent even to nomads in temperate zones. At the moment we do not know a sufficient number of detailed and 111

Although not easy, it is still possible to differentiate between fodder and water in a theoretical analysis. There is no vegetation without water, but where there is water there is not necessarily vegetation. Moreover, not all vegetation is suitable as fodder for livestock and different species of animals Iced on different plants. The amount of water plants need often docs not coincide with the amount required by the livestock feeding off them.

39

Nomadism

as a distinct form of food-producing

economy

precise facts about the dietary systems of the overwhelming majority of nomads (maybe it is anyway too late for specialists to collect these facts), and we know less than we should about the calorie-value of the different elements in their diets. All of this seriously complicates any comparative study of these systems, although the differences between them are too striking to pass unnoticed. The main types of pastoral nomadism The main geographical types of pastoral nomadism may be distinguished according to the criteria so far examined. I fully realise that in making this statement I am supplying those who disapprove of classifications with yet another weapon with which to attack me. Essentially, within the huge expanse of deserts, semi-deserts and steppes which stretches from North China almost to central Europe and from South Siberia to South Africa there are no precise and undisputed borders. Moreover, in the majority of the regions in which nomadism is widespread it is not completely homogeneous. In no way do the nomads within these regions constitute identical groups; owing to specific local adaptation, historical factors and a number of other reasons, some nomads are more similar to the nomads of other regions than they are to their own immediate neighbours. Between these regions in which nomadism is widespread there are intermediate and marginal areas, and it is difficult to classify the pastoral nomadism of these areas as any one of the main types. Finally, each of the main types of nomadism can in turn be subdivided almost ad infinitum into sub-types, forms and subvarieties, etc. I shall assist my potential critics by citing from now on examples of ways in which actual reality cannot always be accommodated in generalizations. Nevertheless, the existence of different types of nomadism, the representatives of each type of which are closer as a whole to one another than to representatives of other types, is an objective fact. Thus the way in which Kroeber (1947:323) only singled out 'the pastoral belt' of Asia is clearly inadequate. Other scholars have already reached my conclusion, although the types of nomadism which they single out rarely coincide (cf., for example, Narol, 1950; Patai, 1951; Patai, 1978; Bacon, 1954; Berque, 1959:481-2; Rathjens, 1969; Johnson, 1969, 13-15; Spooner, 1973:6-8; Coon, 1976:191f.; Khazanov, 1975:6-7). As always, the criterion of classification depends on the aims of the research in hand. In this chapter the main types of pastoral nomadism, apart from the geographical, are singled out on the basis of ecological-economic criteria: species-composition of herds, position in the ecological zone, the nature of nomadic movements and dietary systems. But in the next chapter I shall try to show that the formation of these different types in certain ways was linked also to historical factors. 40

The main types of pastoral

nomadism

North Eurasian type. In this work North Eurasia is understood to comprise the zone of polar deserts and the tundra zone to the south, which stretches from the shores of the Varanger Fjord to 60° north on the banks of the Okhotsk Sea; in the south it reaches the June isotherm +10 centigrade (Berg, 1955:11-16; Milkov, 1977:26-53). Sometimes the forest-tundra zone is also included in North Eurasia. By virtue of the fact that nomadic-reindeer herding is the only fully monospecialized form of nomadism it is also the most homogeneous. It is true that three sub-types may be singled out in the North Eurasian type: Lapp, Komi-Nentsy, and Chukchi-Koriaks, but the differences between them are ethno-cultural rather than economic. Although from the geographical point of view this type is the most isolated one, at least from the point of view of species-composition of herds it is not completely isolated. In Tuva and the surrounding regions and in North Manchuria quite recently non-nomadic reindeer-herders were neighbours of steppe dwellers herding sheep and horses (Petri, 1929; Tugarinov, 1926:76ff.; Vainshtein, 1972:9, 88-99; Vainshtein, 1980:49-50). Admittedly, reindeer-herding has played a secondary role in their economy, while hunting has been more important. The herding of deer in itself is not necessarily linked with nomadic reindeer-herding. To a limited degree reindeer have been herded and maintained by the inhabitants of the taiga and tundra - the forest Nentsy, Selkups, Nganasans, some of the Ugrians of the Ob area, Kets, the forest Entsy, certain groups of Yakuts, Evenks (Tungus), Evens (Lamuts), Oroks, Negidals of the upper Amgun, Tofalars (Karagasy), certain groups of the Tuvinians - who all live basically by a food-extracting economy (Prokofev, 1928:99-100; Popov, 1948:55-68; Nelleman, 1961:91-2; Narody Siberii, 1956; Vasilev, 1962:67ff.; Alekseenko, 1967:65-70; Vasilevich, 1969: 72-9; Obshchestvennyi stroi u narodov Severnoi Sibiri. XVII-nachalo XX v. 1970:40; Vainshtein, 1972:9; Vainshtein, 1980:49-50; Khomich, 1976:1312; Gurvich, 1977:51). However, the number of domestic reindeer kept by these people was very small. Amongst the Khants of the taiga there were 25-30 head per household; amongst the Selkups 1-20 per household; only northern groups of the Selkups, inhabiting the Taz basin, maintained 200-300 head per household; amongst the Evenks it was up to 25 head; amongst the Kets up to 10 head (a herd of 100 being exceptional); amongst groups of Yakuts in the taiga 8-30 head; and amongst the Oroks up to 20 head. Only amongst certain groups in the tundra did reindeer-herding become the basic occupation, and their herds have been counted in hundreds and thousands. Thus, with full justification, we can call nomadic reindeerherding a phenomenon of the tundra. Ecosystems characteristic of the North, in which there are a small number ol specialized forms occupying relatively large niches (Paine, 1971:162), and 41

Nomadism

as a distinct form of food-producing

economy

the fact that agriculture here is unfeasible has meant that the majority of specialized reindeer-herders in the tundra have been pure nomads. The distinct instability of monospecialized nomadic reindeer-herding leads to a situation in which maximization of numbers in the herd becomes economic strategy, as also in a wider context does striving towards supplementary sources of subsistence. However, this maximization has nothing in common with what Barth has suggested of nomads in the Middle East; its aim is not to accumulate livestock as a form of capital, but it is first and foremost to preserve a traditional economy in a specific ecological situation. Secondary forms of economic activity, which are characteristic of all nomads (including pure nomads) in the North were connected with food-extracting (hunting and, most particularly, fishing). Pastoral seminomadism is a well known feature of the North Eurasian type of nomadism, but it is semi-nomadism of a very specific form comprising a more or less organic combination of reindeer-herding with a food-extracting economy. For example, there are well known groups of European Nentsy and Chukchi who live by a mixed economy, combining the pasturing of reindeer (on average 100-150 head per household) with fishing and maritime hunting (Maslov, 1935:111-15; Koviazin, 1936:3; Narody Sibiri, 1956:902). It is interesting that as a whole occupation with hunting and fishing, like occupation with agriculture, has limited the mobility and pastoral migrations of the aforementioned groups, which have been shorter, both temporally and spatially, than those of pure reindeer-herders. Nevertheless, reindeer-herding has not known the diversity of forms of pastoral economy changing one into the other, which are inherent in other types of nomadism. The ecological character of the tundra is such that reindeer-herding occupies the dominant position there. Increase in the number of domestic deer leads to a decline in the number of wild ones and to corresponding difficulties for hunting groups (Gurvich, 1977:52). As reindeer-herding developed wild reindeer were gradually forced out of different areas of the tundra and now they are only to be found in reasonably large numbers on the Taimyr peninsula and by the lower reaches of the Indigirka and Kolyma rivers; in these regions the hunting of reindeer was the basic occupation of the indigenous population (Druri, 1949:25; Krupnik, 1976a:65; Simchenko, 1976:4), but even here the number of wild" reindeer is rapidly diminishing. Consequently, in this case we cannot talk of a combined utilization of one zone. The economy of the nomadic reindeer-herders disrupted the ecological foundation of the economy of the hunters. Essentially, reindeer-herding could have occupied not only the dominant, but even the monopolizing position in the tundra. However, things did not work out in such a way (Sermerikov, 1933:9-11). Leaving aside the groups and individuals who came from agricultural societies in the south, the nomads divided up the tundra between themselves and the groups which 42

The main types of pastoral

nomadism

primarily were involved in hunting and fishing and practised reindeerbreeding only on a small scale, i.e. those inhabiting the tundra between the Yenisei and Kolyma rivers - the Nganasans, the tundra Entsy, certain groups of northern Yakuts, and the Yukagirs. The number of reindeer herded by these peoples was greater than the number kept by the taiga hunters, but far smaller than the number kept by the nomadic reindeerherders. Thus, in poor Nganasan households there were no more than 50 head of deer (Popov, 1948:55), in poor households of the tundra Yukagirs up to 10-20 head and in rich ones 100-150 head (Gurvich, 1975:58), and in poor northern Yakut households there were up to 20 head and in rich ones between several hundred and a thousand head (Gurvich, 1977:50). From an ecological point of view the co-existence of these groups with nomadic reindeer-herders has been based on a system of enclaves, rather than of niches. Niches are not necessarily linked with the spatial division of a particular ecological zone, whereas enclaves are. In the present context an enclave signifies a definite and compact territory within an ecological zone, the economic utilization of which is essentially specific in comparison with the utilization of analogous territories. In addition, the groups utilizing these enclaves occupy the monopolizing position within them. The pastoral migrations of nomads in North Eurasia have been timed according to clear-cut seasonal changes, thus they are stable, linear and meridional. Availability of fodder is the most important criterion in the choice of both winter and summer pastures. Other criteria which have affected choice of summer pastures are possibilities for finding further, supplementary forms of economic activity and the location of blood-sucking insects. Nomads in North Eurasia used to spend 4-5 months in winter pastures, 2 months in summer pastures and 5-6 months travelling between the two. When the herds were actually pasturing migrations took place between once and three times a month, but in the months when they were between pastures the herds were moved every day, except during fawning (Kertselli, 1911:5ff.; Karev, 1968:32-3; Krupnik, 1976:66ff.). Length of pastoral migrations was limited to the width of the tundra zone, i.e. the land between the sea and northern border of the taiga; distances varied from 50-100 kilometres amongst the reindeer-herders in Chukotka, 100-200 kilometres amongst the Lapps in Finmark, 150-200 kilometres amongst the Nentsy of the Kanin-Timan tundra, 320-400 kilometres amongst other groups of Lapps, 400-500 kilometres amongst the Chukchi of the Elvuney and Anyuy areas to 1200 kilometres amongst certain groups of the Nentsy in the great Tundra (Gondatti, 1897:171; Semerikov, 1933:6-8; Terletsky, 1934:43; Vorren and Manker, 1962:20; Paine, 1972:76; Krupnik, 1976:66). The main feature of the dietary systems of nomadic reindeer-herders has been the emphasis on meat consumption. Blood also features. The products of reindeer-herding have been supplemented with fish and fowl, but only 43

Nomadism

as a distinct form of food-producing

economy

some groups in the tundra milk their reindeer (Nelleman, 1961:92). For example, the Nentsy knew nothing of milking (Khomich, 1976:75), but according to some field research they sometimes sucked milk from their does. In the opinion of Spooner (1973:5) the reindeer-herders of the North were the only nomads who used no agricultural products; Simchenko (1976: 83-4, 284) insists that those groups which eat reindeer meat, whether of their domestic or of wild deer, have no need of vegetables in their diet. It would appear that neither assertion is entirely correct. The Lapps, European Nentsy and later on the Chukchi procured agricultural products by way of exchange. However, it must be remembered that before large herds became a feature of reindeer-herding in the North regular exchange operations with southern societies had been established. Already at the beginning of the eighteenth century the European Nentsy could not do without Russian flour (Kolycheva, 1956:86). Furthermore, account must also be taken of gathering which was very widespread amongst reindeer-herders. 'Linnaeus' idea that vegetable foods played a quite subordinate part in the Lapps' diet is probably not entirely in accordance with the facts, judging from other sources . . . a well-known mountain plant, Angelica archangelica . . . has been used for centuries as a vegetable by the Lapps and . . . probably formed an important element in large parts of the diet of subarctic peoples' (Fjellstrom, 1964:100; see also Georgi, 1776:8-9; Beretti, 1929:23; Menovshchikov, 1974:93-4 on the use of vegetable food amongst the Chukchi and Koriaks). The only thing it is safe to assert here is that vegetable foods have played a much smaller part in the diet of reindeer-herders of the North than they have in the diets of other nomads. Eurasian steppe type. From the geographical point of view this type occupies (occupied, more precisely) the huge zone of steppes, semi-deserts and deserts of the temperate zone which stretch from the Danube (the Hungarian puszta) to North China. For the sake of brevity and in accordance with the leading economic role played by the steppe landscape in the lives of the majority of nomads in this region I refer to it as the Eurasian steppe. In the south where the deserts of the temperate zone become the deserts of the subtropical zone (the so-called Iranian-Turanian or South Turanian deserts, Petrov, 1964:51; Petrov, 1966:16), the Eurasian steppe type of nomadism becomes Middle Eastern. The entire temperate zone of Eurasia subdivides more or less neatly into areas favourable for agriculture, areas favourable for extensive pastoralism and marginal areas in which both these forms of economic activity are possible. In the latter areas the predominant economic activity depends on specific historical circumstances and is not determined directly by ecology (Khazanov, 1973; Khazanov, 1975a; Khazanov, 1978b). In the wooded steppe agriculture which does not require irrigation is 44

The main types of pastoral

nomadism

possible, but in the open steppe agriculture is threatened by drought for between 3½and 6½years out of every 10 (Fedorovich, 1973:210-11). It is not surprising that 89½% of Mongolia's territory is given over to fodder crops, out of which only1½%has to be given over to hayfields (Kalinina, 1974:143). Nomads have been able to use the wooded steppe only in summer on account of the fact that in winter it is covered with thick snow. But then in the semi-deserts and deserts only agriculture which requires irrigation is possible, although in certain territories which mainly are to be found on the banks of rivers and inland reservoirs, and also in foothills and hollows between mountains some exceptions may be observed. Although over the course of several thousand years the Eurasian steppes have been one of the main regions of nomadism in the world, one fact is often forgotten. This is that in some historical periods political factors determined the predominance of pure nomadism, but in others seminomadism was fairly widespread." In the Kazakh steppes and the lands of the Golden Horde to the north of the Black Sea, where the Mongol invasion caused so much damage, agriculture nevertheless did not completely disappear (see data presented by al-Umari and al-Ayni in Tizenghauzen, 1884:230, 233, 513; Rubruck in The Mongol Mission, 1955:100; Josapha Barbaro In Travels to Tana and Peisia, 1873:21; Mekhovsky in Mekhovsky, 1936:59; see also Zajaczkowski, 1968:229-33). Earlier, in the ninth century, Tamin ibn Bahr mentions 'settlements and cultivated lands' in the land of the Kimaks (Minorsky, 1948:284; Kumekov, 1972:94-5). There are records of an agricultural sector in the economy of the Oghuz (Karryev, Moshkova, Nasonov, Jakubovsky, 1954:41-2; Agadzhanov, 1969:260). Judging from the written sources, even the Hsiung-nu to some extent'. . . were occupied with the cultivation of the land'(Taskin, 1968:24,121). Of the Sarmatian tribes Strabo (XI.2.1) writes: '. . . some [are] . . . nomads and others tent-dwellers and farmers'. There was also an agricultural sector in the economy of the Scythians (Khazanov, 1978: 429-30), the Wu-sun (Akishev, 1970: 69) and the ancient nomads of the Altai (Griaznov, 1947:14-15). The sedentarization process was revitalized for the last time from the end of the eighteenth century, as gradually most nomads were conquered and drawn into the Russian empire. The enforced sedentarization of the thirties of this century completed this process in the USSR. However, already towards the beginning of the twentieth century in Kazakhstan, for example, pure nomads were only to be found on the Mangyshlak peninsula, along the Syr-Darya river and in the desert part of Central Kazakhstan. Coon (1976:224) has quite rightly pointed out, as Bacon (1954:46) has 11

l lowever, my once asserted opinion that in the course of most of the history of the nomads of the Eurasian steppes semi-nomads have tended to predominate numerically over nomads (Khiuinnov, 1975:11) now, it seems to me, needs to be tested further.

45

Nomadism

as a distinct form of food-producing

economy

not, that in nomadism of the Eurasian steppe type the dividing line between nomad and agriculturalist is much less distinct than it is in the Near Eastern type. Although the boundary between pure nomadism and semi-nomadism, also between those and other forms of extensive pastoralism, was much less distinct and easier to cross in the Eurasian steppes than it was, for example, in Arabia, on the other hand it was also much easier to cross back over it. Correspondingly, different transitional forms were more numerous here, but they were also less fixed. The species-composition of herds and even the percentage correlation of different species of livestock in the same territories of the Eurasian steppes, on the whole, have been stable over the entire period during which nomadism has existed there. Apposite examples here are the ancient Sarmatians and the Kalmucks of the modern period (Khazanov, 1974), the inhabitants of North Karakalpakia in the eleventh to the fourteenth centuries and in the most recent past (Tsalkin, 1966:154), the nomads of Tuva in the first millennium A.D. and in the nineteenth century and the beginning of the twentieth (Vainshtein, 1980:52ff.), and the Mongols of the Middle Ages and their descendants in the present day (Tsalkin, 1968:22-3). The attempt of Eberhard (1952:107-39; cf. Johnson, 1969:13) to divide the Eurasian steppe type of nomadism into the Mongol, in which horses predominate, and the Turkic, in which the species-composition of herds is mixed, has no foundation in reality. Mongols herded animals of different species, and some Turkic-speaking nomads owned as many horses as Mongols did. However, the Eurasian steppe type of nomadism can of course be divided into several sub-types: the Inner Asian (Mongol), the KazakhstanianMiddle Asian, the East European and the South Middle Asian (Turkmenian). It should be pointed out that the differences between these sub-types are ethno-cultural, rather than economic. However, the South Middle Asian sub-type rather stands out on its own, for in many respects it is similar to the nomadism of the Middle East and thereby even to the Near Eastern type. The specificity of the species-composition of herds in the Eurasian type of nomadism is determined by the leading role of two species: sheep and horses. The value of sheep scarcely needs any comment. The variety of plants which sheep eat is very wide and includes many different kinds of plants which are unfit for horses and large stock. Also important is the fact that sheep can get at fodder in pasture covered with snow up to 15-17 centimetres deep (Nasimovich, 1955:75). In contrast to attitudes in other regions, in the Eurasian steppes sheep have been esteemed animals from the earliest times. According to Uzbek belief sheep came on earth straight from heaven (Narody Srednei Azii, 1962:221). 46

The main types of pastoral

nomadism

In the steppe the horse is as valuable as the camel is in the desert. In the words of the Kazakh Khan Kasim: 'We are men of the desert, and here there is nothing in the way of riches or formalities. Our most costly possessions are our horses, our favourite food their flesh, our most enjoyable drink their milk and the products of it. In our country are no gardens or buildings. Our chief recreation is inspecting our herds' (Mirza Muhammad Haidar Dughlat, 1895:276). The horse was most important as a transport animal (particularly for riding), and also as a source of meat and milk. According to a Kazakh proverb it is not the man who is reproved who dies, but the man who loses his horse. On the whole the importance of goats has become secondary to that of sheep, although in certain regions (particularly desert regions) goats have become more important. As is the case in other regions, the goat in the Eurasian steppes is known to be a somewhat inferior equivalent of the sheep. In Mongolia there are very many goats in the dry and rocky regions of the desert steppe of the Gobi and the mountainous steppe of the Altai (Shulzhenko, 1954:38). When, as a result of the Chinese colonization of Inner Mongolia, the nomads there were forced to use poorer pastures, goats, rather than sheep, began to occupy a more important place in the nomads' herds (Lattimore, 1967:74n.23). At the beginning of the nineteenth century one of the Russian travellers made the following observation about the Mongols: 'Goats are the lot of those with insufficient means' (Timkovsky, 1824:79; cf. Shaniiazov, 1975:189 on the semi-nomadic Uzbeks). The importance of large stock has become secondary to that of small stock, nevertheless the former have occupied a more important place in the composition of herds in Eurasia than they have in the Near and Middle East. The specific significance of large stock in the composition of herds has followed two rules: in steppe regions there were more large stock than there were in desert regions; and the greater the number of large stock, the more important the role of agriculture in the general balance of the economy. For the mobile Eurasian nomads who did not lay in fodder for future use the cow was admittedly a valuable animal, but it was too capricious and ill suited to being driven over long distances. Rychkov (1877:22) wrote that cows '. . . cannot pasture in the steppes, for this reason the Kirghiz [Kazakhs - A. K] keep few of them, and the richest man never has more than twenty.' In a twenty-four hour period a cow will eat no more than 48 kilograms of grass; but if conditions are not ideal (the grass is too short, too tall, or rare, or the relief of the pastures is difficult) it will eat less and have to remain hungry, for it has no way of increasing its feeding time and the number of mouthfuls it can tear up for itself to eat in a day is limited (Ficlstrup, 1927:83; Baskin, 1976:114). Consequently, large stock appeared amongst the Kazakhs only in the 47

Nomadism

as a distinct form of food-producing

economy

second half of the eighteenth century and the beginning of the nineteenth; this was the period in which Russian power began to restrict the freedom of the Kazakhs' pastoral migrations and revitalized the process of their sedentarization. Some large stock was appropriated from the Karakalpaks, some was just taken from the Kalmucks at the time of their unfortunate flight from Russian power in the lands round the Volga into Jungaria (Georgi, 1776:126, 127; Levshin, 1832, part iii: 189). Amongst the Adai Kazakhs, even in 1927 cattle constituted only 4% of the total livestock; moreover, most of it belonged to that 3 % of the Adai which had already been sedentarized (Briskin, 1929:39). According to the data for 1927, in Kirghizia large stock comprised between 34.3% and 46.5% of the total number of livestock in Russian households and between 40% and 47.3% in Uzbek ones; both types of household consisted of sedentary agriculturalists. Amongst the Kirghiz who were still leading a nomadic and semi-nomadic way of life large stock comprised between 6.85% and 10.3% of the total number of livestock, and amongst those already seriously involved with agriculture between 14.8% and 24.1% (Lus and Kolesnik, 1930:46). On the other hand, the semi-sedentary Karakalpaks in the nineteenth century kept more large stock than they did small stock (Narody Srednei Azii, 1962:439). An increase in the number of large stock has also been registered amongst other nomads when they began to lead a more sedentary way of life (see, for example, Karmysheva, 1954:111 ontheLokai Uzbeks or Erdniev, 1964:25 on the Kalmucks). True, in antiquity nomads used oxen as transport animals; but in the Middle Ages, particularly in the post-Mongol period, oxen gradually ceased to be used as the mobility of nomads generally increased. In the last millennium B.C. ox-drawn kibitkas were characteristic of the Scythians (Pseudo-Hippocrates, De Here, 25) who roamed the Eurasian steppes. In the nineteenth century these kibitkas were used only by the Karakalpaks (Narody Srednei Azii, 1962:441), the Kalmucks (Zhitetsky, 1893:37), the Nogai (P. P. Ivanov, 1958:43; Shennikov, 1973:53) and by the Chumaks, Ukranian peasants, who specialized in transporting salt from the Crimea. It is very rare to find fossil remains of camels from the last two thousand years B.C. in the Eurasian steppes. Ssu-Ma Ch'ien, describing the pastoralism of the Hsiung-nu, maintains that camels are rare animals (in Taskin, 1968:34). It would appear that the camel became more important for the nomads of the Eurasian steppes in the Mongol and post-Mongol periods. It has been maintained that Mongols began to herd camels on a relatively large scale after they had seized herds of camels from the Tangut at the beginning of the thirteenth century (Vladimirtsov, 1934:36; Kychanov, 1977:47, 50). The cold and damp climate in the Eurasian steppes has meant that camels, 48

The main types of pastoral

nomadism

incapable as they are of independently getting at fodder in pastures covered with snow, are not widespread here. In winter the Kazakhs used to cover their camels with horsecloths to protect them from the cold (Georgi, 1776:127; Fielstrup, 1929:90). It was only in the southern, desert regions, mainly amongst the Turkmen and certain groups of Kazakhs and Mongols, that the camel was of essential importance. In the south of Mongolia with its dry, desert pastures the camel has almost completely taken the place of large stock and is even successfully competing with the horse (Shulzhenko, 1954:38). The Kazakhs and Kirghiz for the most part have herded Bactrian or hybrid camels and the Turkmen have herded dromedaries (Ishchenko et ai, 1928:135; Argynbaev, 1973:156). In contrast to the Near East, in the Eurasian steppes the basic use of the camel has been as a means of transport, only relatively rarely has it been used here as a riding animal. However, camels have also been used for milking. This author has himself been able to savour the delights of Turkmenian chal. In summer camel-milk was one of the staple foods of the Adai Kazakhs (Ishchenko et al., 1928:146). The donkey has never played an important role in the lives of the nomads and semi-nomads of the Eurasian steppes. According to a Turkmenian proverb the wealth of a poor man is a donkey and a mangy billy-goat (Babadzhanov, 1975:222). Another characteristic of the Eurasian steppe type of nomadism was that in the composition of herds small stock was always combined with those animals used for riding and transportation. However, those animals had a polyfunctional purpose. Bacon (1954:46) thus is being fairly accurate when she describes this type of nomadism as 'multi-animal'. In the Eurasian steppes the tendency to specialize in the breeding of animals of only one species has existed, but it has not been widespread. Thus, in South Tadjikistan the Lokai have specialized in the breeding of horses, and the Qarluqs in sheep. About the latter there even exists a particular saying: God made the Qarluq for sheep (Karmysheva, 1954:64ff.). Of course, a mixed species-composition of herd does not mean that animals of different species are necessarily pastured together; however, it has not been uncommon (particularly in winter) for different species to be pastured on a rotational basis on the same land. What a mixed herd does mean is that each species in it is necessarily part of one household. In contrast to Arabia and the Sahara, in the Eurasian steppes there were no pure camel- or horse-herders, nor were there any pure sheep-breeders. To a considerable degree this is explained by the fact that Eurasian steppe nomadism existed in a temperate climatic zone, and also by the wintertime problem of fodder which the livestock had to get from underneath snow. One solution to this problem has been provided by the horse with its unique 12

11

chat - f c r m e n l M l aimd's-milk.

49

Nomadism

as a distinct form of food-producing

economy

ability, when pasturing in snow, to uncover grass up to 10 centimetres deep with just three kicks. The horse can pasture grass covered with 30-40 centimetres of snow, sometimes even 50 centimetres (Nasimovich, 1955:74; Baskin, 1976:55). In the steppe itself, both the Mongol steppe and that part of the steppe to the west of Mongolia, which in the Middle Ages Arab and Persian authors called Dasht-i-Qipchaq, and even in the East European steppe, which medieval Russian chroniclers called Dikoe Pole (Wild Field), nomads occupied an all but dominant position. Agriculture along the banks of rivers and reservoirs has taken up so very little of the wide open steppes that we can hardly pretend there has been a joint season-by-season utilization of one ecological zone here. All we can say is that more often than not agri­ culturalists have occupied certain enclaves while nomads have occupied other ones. From Avesta to Shah-nama 'Iran' (i.e. the sedentary world) was a contrast to 'Turan' (i.e. the nomadic world) not only from an economic, but also from a territorial point of view. With time the ethnic implications of these words changed, although the contrast itself remained unchanged. The position was somewhat different in marginal areas: the Hungarian plain; Maveraunnahr, the region between the Amu-Darya and Syr-Darya; Semirechye; Ordos, and others. In some of them an agricultural culture stretches far back into ancient times. It was not rare for nomads and semi-nomads, on the one hand, and agriculturalists, on the other, to live side by side in these areas. There was joint utilization of the same zones and niches here and almost always it was accompanied by clashes and fighting. The fact that the Eurasian steppe type of nomadism for the most part was situated in a temperate zone was one of the most important in determining the routes and character of pastoral migrations. Except in mountainous regions where vertical migrations either had to take place all year round or else they were seasonal (such migrations were particularly widespread amongst the Kirghiz, certain groups of Mongols, Tuvinians and Uzbeks), migrations were usually regular, linear, meridional, routes were fairly stable, seasonal changes of pasture were clear-cut and water was the main priority in summer, fodder in the remaining seasons. 'The fodder requirements of the herd, the necessity of providing it with water and the best way of protecting it from the cold in winter are Jhe basic factors in the economic life of the nomad. The amount of time spent in camps, the direction of migrations, number of pastoral migrations and distance of the latter all depend on those factors' (Ishchenko et al, 1928:105). This observation about the Kazakhs can be applied to almost all the other nomads in the Eurasian steppes. Only in certain desert regions, for example amongst the Turkmen, were there radial-circular migrations in which availability of water was the primary concern (Orazov, 1975:217). The regular character of migrations was determined by the clear-cut 50

The main types of pastoral

nomadism

seasonal changes of the temperate zone. 'For a nomadic household migrating many hundreds of kilometres from north to south, it is very important for there to be fodder crops along the entire migration route at a stage when they are nutritionally valuable. The arrival of a nomad in July in the North with his livestock still scraggy after winter is tantamount to jute [a huge loss of stock] because the grass begins to dry out at this period and is almost two times less nutritious than it is in May and June' (Ishchenko et al., 1928:44). The same circumstances have determined the meridional character of pastoral migrations which was established in ancient times and changed little in the Middle Ages, at least in relatively peaceful periods. Owing to the meridional character of their migrations, even the khoshuns, administrative units in Mongolia before the Second World War, were, as a rule, oval in form, stretching from north to south (Graivoronsky, 1979:47). Nomads were drawn to the North by the fodder rich summer pastures there, for in the summer the vegetation of the steppe does not usually stop growing (Alekhin, 1950:257), and to the southern winter pastures because the winter is relatively warm and the snow is not too deep, although there has often been insufficient fodder in these pastures. The nutrition value of fodder stocks in desert pastures becomes 2-2£ times lower in the winter (Sredniaia Aziia, 1969:174). The summer pastures where the fattening of livestock took place have been particularly important for the majority of nomads. In the steppe zone of Middle Asia 1 sheep needs 1 hectare of pasture, in the desert a sheep needs 3-6 hectares, and in some places even 10 (Sredniaia Aziia, 1969:174). In the nineteenth century when the majority of summer pastures used by the Kazakhs were seized by the tsarist government, the crisis of the nomadic economy of the Kazakhs began. The permanency of routes of pastoral migrations can often be traced over a given territory from antiquity up to the modern period. Strabo (Strabo, VII. 3.17) wrote of the nomads of the Eurasian steppes in antiquity '. . . they follow the grazing herds, from time to time moving to other places that have grass, living only in the marsh-meadows about Lake Maeotis in winter, but also in the plains in summer'. Piano Carpini (The Mongol Mission, 1955:55) makes a similar assertion about the nomads of the Golden Horde: 'All these men go south in the winter towards the sea, and in the summer they go up north along the banks of the same rivers to the mountain.' Observers of the end of the nineteenth century and the beginning of the twentieth also recorded that the Kalmucks left the open steppe for winter camps in estuaries and lakes in the Manych valley and adjacent areas overgrown with reeds (Dubrova, 1898:187; Ochirov, 1925:15). Thus, for over two thousand years the routes of pastoral migrations in East Europe have remained virtually unchanged. 51

Nomadism

as a distinct form of food-producing

economy

Different observers repeatedly have emphasized the stability and fixedness of these routes. Nebolsin (1852:26) wrote of the Kalmucks: 'Once one has comprehended their system of pastoral migration inculcated by time and habits, which only changes on account of particular topographical conditions and a decrease in places suitable for migrations, one can almost always know for certain where in particular a clan, or a certain part of this clan, have their kibitkas in the steppe.' Of the Kazakhs it has been recorded that'. . . to go along the route which another aul uses is considered to be a crime' (Lialina, 1894:30; see also Levshin, 1832, m:13). Consequently, the distance of pastoral migrations has varied very considerably. Amongst the Mongols of Inner Mongolia it sometimes has been less than 150 kilometres (Lattimore, 1967:73n.21), in the Gobi zone of Outer Mongolia it has amounted to 600 kilometres (Graivoronsky, 1979:49) and amongst the Kazakhs of the Little and the Middle Hordes it has been 1,000-1,500 kilometres {Narody Srednei Azii, 1963:354). Amongst the Kirghiz the extent of vertical pastoral migrations has varied between a few dozen and 150-200 kilometres {Narody Srednei Azii, 1963:354) and amongst the Turkmen the extent of radial-circular migrations has varied between 20-30 and 150-200 kilometres (Orazov, 1975:216-17). The dietary systems of nomads of the Eurasian steppe type were connected with their location in the temperate climatic zone. They used less meat than the reindeer-herders of the North or the pastoralists of High Inner Asia, but more than the nomads of the Near and Middle East and all the more of East Africa. It is possible then that nomads of the Near and Middle East used more vegetable foods, but I cannot verify this assertion with any precise data. According to the calculations of Maisky (1921:156), in the calorie intake of Mongols at the beginning of the twentieth century milk products accounted for 55.31%, vegetable foods for 24.38% and meat for only 20.31%. In the mid fifties of the twentieth century the nomads of Mongolia on average per head consumed 400 grams of meat a day or 143 kilograms a year (from 128 kilograms in poor households to 255 kilograms in rich ones), 0.78 litres of milk a day or 260 litres a year (from 107 litres in poor households to 546 litres in rich ones) and other milk products, for which 0.68 litres of milk a day or 248 litres a year were needed; in addition, 18 kilograms of flour a year, or 49 grams a day, were consumed (Zhagvaral, 1974:115-16). As a whole the dietary systems of the nomads of the Eurasian steppe type have centred on two basic elements, meat and milk, with an obligatory supplement of vegetable foods. The assertion of Bacon (1954:46) that they can exist for an unlimited period on only milk products and meat is plainly 13

13

There is data for the consumption of flour for 1948-9. In these years there was a serious deficit in Mongolia of agricultural products, so the latter had to be supplemented with the products of gathering. The usual consumption of flour was much higher.

52

The main types of pastoral

nomadism

false. In the eighteenth and nineteenth centuries the most severe punishment the Khans of Khiva could inflict on the recalcitrant Turkmen was to forbid the latter to use the bazaars where they could buy the grain they needed (Narody Srednei Azii, 1962:97). The ratio of meat and milk products in the diet has varied according to season. In the summer milk products have predominated, in the winter meat (Zhukovskaia, 1979). The meat most commonly eaten has been mutton, and goat's-meat where goats were herded; but horse meat has also been used as food, and this fact sharply distinguishes Eurasian nomads from all other nomads. In compar­ ison to the meat of other animals bred in the Eurasian steppes, horse meat has a higher calorie value, a higher percentage of mineral substances, amino acids, vitamins A and B, etc. (Gunga, 1976:31). On the other hand, camel meat has not been an important food. Abu Dulaf, a tenth century author, wrote that the Kirghiz ate all meat except for camel. All animals herded, including sheep, have been milked. Testimonies of the way in which the nomads of the Eurasian steppes have used blood as food stretch back to antiquity. It may be assumed that in antiquity and the Middle Ages (see, for example, the sixteenth-century Polish author Mekhovsky, 1936:59) blood was a more significant part of the diet than it has been in the modern period when the prohibitions of Islam have held sway. The Mongols use blood as food to this day (Zhukovskaia, 1979). Nevertheless, blood has never played an important role in the dietary systems of the nomads of the Eurasian steppe type. The Near Eastern type. The Near Eastern type of nomadism is significantly less uniform than the types of nomadism examined above. It can be divided precisely not only into four basic geographical sub-types - the Arabian, North African, Saharan and Northeast African - but, if so desired, into a great many more. The above mentioned sub-types are themselves very diversified. Moreover, the distribution of different forms of extensive pastoralism and even of different variants of pastoral nomadism itself very often does not correspond to standard geographical borders. Nevertheless, despite historical considerations which are of no small importance, a number of the general economic characteristics which nomads of this region share enable us to recognize that there does exist a Near Eastern type of nomadism. In the Near East all the mentioned forms of extensive pastoralism are present, but with local particularities according to their distribution. The real state of affairs can be simplified with the statement that Arabia and the Sahara are the domain of pure nomads, while North Africa, Palestine and partly Syria and Iraq are the domain of semi-nomads and those involved in herdsman husbandry. According to a well-known expression, Iraq is the grave of Bedouins. But this could be put another way: out of pure nomads 53

Nomadism

as a distinct form of food-producing

economy

Iraq made semi-nomads (for a description of the latter see Oppenheim, 1939:22ff; cf. Fernea, 1970:11). At the same time the economic boundaries between the different forms of pastoralism in the Near East, most particularly between pure nomads and the other pastoralists, appear to be more clear-cut than they are in Eurasia. The fact that pure Bedouins are so different from the other pastoralists and agriculturalists of the region, in the final analysis, was brought about by ecological specificity, but the difference was consolidated by economics, social and political practice and their system of values. Only in North Africa where shifting plough cultivation is possible (Nicolaisen, 1963:183) do the economic borders between different forms of pastoralism seem less clear-cut, more passable and reversible, than they are in Arabia and the Sahara. However, the nomads themselves in Arabia may be divided up into pure camel-herders as, for example, the Rwala, the Shammar and the Mutair, and sheep- and goat-herders who also keep some camels (mainly for transport needs), full-Bedouins and half-Bedouins (Dickson, 1951:109). But this division should not be overestimated as, for example, the Mutair, basically a camel-herding tribe, also keep small herds of sheep and goats. In some South Arabian tribes there are fewer camels than there are in North Arabian ones, more goats, and the donkey is the basic riding and transport animal (Bird, 1834:196; Coon, 1959:872). Sheep-herders have lived mainly in Najd, Qasim and Jabal Shammar, and also in Syria, Palestine and North Africa. Many, if not the majority of them, essentially have been semi-nomads (Burckhardt, 1830:330-1; Wallin, 1850:35; Doughty, 1888, i:15, 45, 234, 440; for a summary of data see Pershits, 1961:36-9). The three most important animals determining the species-composition of herds in the nomadism of the Near Eastern type are the camel, sheep and goat. Volney (1959 [first published 1787]:203) long ago noted: 'Telle est l'importance du chameau pour le desert, que si on Ten retirait, on en sous trairait toute la population, dont il est l'unique pivot.' Almost everywhere the camel is the most important riding and transport animal; it is also important for the milk it gives. As is the case everywhere else, goats for the most part take the place of sheep in the most adverse or mountainous condi­ tions, although ethno-cultural traditions clearly also play a specific role here. Large stock, if we ignore certain specifically local and furthermore not purely nomadic groups, are of essential importance only in the Northeast African sub-type. Despite the high prestige value and partial military significance of the horse, everywhere in the Near East there are fewer horses than there are camels in the Eurasian steppe type of nomadism. According to the data of Burckhardt (1831:57), amongst the Anaza at the beginning of 14

u

In Dhofar on the shore of the Indian Ocean, one group herds hump-backed cattle, and in Southern Iraq a group pastures water-buffalo (Coon, 1943:209-17; C o o n , 1959:872; Thesiger, 1964).

54

The main types of pastoral

nomadism

the nineteenth century on average there was only one horse for 6-7 families. 'Horses are of no economic value but serve merely as weapons for procuring booty and influence' (Musil, 1928:371). The horse, basically, was used as a riding animal; it was not pastured with the herds and its meat and milk were not used for food. On the other hand, here the donkey is more important as a transport animal than it is in the Eurasian steppes, but would seem to be less important in this capacity than it is in the Middle East. A donkey can carry up to 100 kilograms, which is only two to three times less than a camel can carry. However, the donkey is still the animal of semi-nomads (the shawiya of the Arabs), the vassal tribes of the Tuareg, other extensive pastoralists and also agriculturalists. Having quite forgotten that donkeys were the basic transport animals if not of their direct ancestors, then of their forerunners in the Bronze Age, true Bedouin nomads and noble Tuareg do not ride donkeys. While I consider the particular role of the camel to be the first important characteristic of the nomadism of the Near Eastern type, the second is the separate pasturing of animals of different species. Camels cannot be pastured with small stock. In the Eurasian steppes sheep and goats have snow in winter instead of water, but in hot deserts, even in winter, these animals (in contrast to camels) have to be watered every 3-4 days. In certain situations - in Arabia and in the Sahara - this leads to intergroup and intertribal specialization, in others - in North Africa - it sometimes leads to specialization within a tribe, and in other situations - in the Horn of Africa - to specialization within a group, or even within a household. It is true that this statement constitutes a generalization made out of many different and specific situations. However, in all areas where the Near Eastern type of nomadism is present, camels are pastured separately from other animals. The laws of ecology in this respect are more immutable than many human laws. This state of things in itself has determined many of the characteristics of the pastoral migrations of the nomads of the Near East. Camel-herders have migrated more and for longer than sheep-herders, they have been less restricted by necessity to remain close to water-sources. In North Africa often only camels have been taken to the winter pastures in the desert, while sheep have remained in Tell (Dhina, 1956:421-4). Julian Tuwim once quipped that a camel can work for a week without drinking, while a man can drink for a week without working. By 'drinking' the Polish poet meant alcoholic drinks. However, if we substitute plain water for alcohol, the aphorism (at least that part of it which refers to the camel) is completely appropriate. In Arabia, the Sahara and partly in the Horn of Africa nomads regularly m i g r a t e to agricultural oases with which they have numerous historical,

Nomadism

as a distinct form of food-producing

economy

ethnic, economic, social and cultural ties. In Arabia, particularly on the borders with Syria and in Qasim, nomads follow an ancient tradition of pasturing their livestock in fields where the crop has been harvested (Pershits, 1961:31; Smilianskaia, 1979:27). But this kind of joint utilization of one ecological zone is less important here than it is in the Middle East, both because the agricultural oases are relatively small and because they are primarily given over to the cultivation of date palms, the waste products of which provide less food for livestock than the stubble of grain-crops does. Only in North Africa and the countries of the Fertile Crescent did this joint utilization of one ecological zone with agriculturalists acquire great significance. One other characteristic of the pastoral migrations arises from the location of the Near East in the hot, dry zone. The irregularity of precipitation in many areas and the indefinite localization of pastures means that the routes of pastoral migrations are also less stable, and in some places less regular than, for example, in the Eurasian steppes and in the Middle East. Capot-Rey (1953:251) remarks of the nomads of the Sahara: 'The first rule of the game - to tell the truth the only rule - is that the nomad follows the rain.' Although pastoral migrations can be seasonal, they are sometimes less regular in this respect than the migrations in the regions already discussed; in effect availability or absence of water is the decisive factor (Marx, 1978:46-9). All these characteristics of pastoral migrations are to be found more conspicuously among the nomads of Arabia and the Sahara, partly also of the Horn of Africa, than among any others. In North Africa the routes of pastoral migrations are much more stable. However, in all areas there are local variations. Some nomads of the Sahara spend the dry period outside the desert - in Tell to the north or in Sahel to the south, others spend the summer in Sahel and return to the Sahara for the winter, while others - the nomads of the ergs - depending on availability of fodder and water, roam in the Sahara itself and, moreover, their pastoral migrations are not obviously seasonal (Capot-Rey, 1953:254-61; cf. Johnson, 1969:116). But in the semi-desert areas to the north of the Sahara, because of the rainy season there, pastoral migrations are more stable (Capot-Rey, 1953:18). Amongst camel-herders in Arabia the localization of winter and summer pastures depends on the rains and is indefinite (Musil, 1927:45; Coon, 1959:873; Johnson, 1969:45). At the same time the pastoral migrations of the Rwala and the Mutair are quite remarkably seasonal, taking place along nonlinear (elliptical) routes. With the exception once again of the North African sub-type, linear routes of migrations do not so completely predominate amongst the nomads of the Near East as they do amongst the nomads of North Eurasia and the Eurasian steppe. In the Near East nonlinear routes are not uncommon. 56

The main types of pastoral

nomadism

Some Arabian tribes combine a meridional and circular route with a vertical one (Pershits, 1961:31). The distance of pastoral migrations varies considerably according to the different conditions of each area. Amongst the Rwala they were as much as 1,500 kilometres (Awad, 1962:333), amongst the Tuareg between 300 and 1,000 kilometres (Capot-Rey, 1953:262), and amongst the Tuareg of Ahaggar no more than 50 kilometres (Nicolaisen, 1963:149). The basic foods of nomads in the Near East are milk and vegetable produce, for here meat is less important than it is in nomadism of the Eurasian type (see, for example, Nicolaisen, 1963:38 on the Tuareg; Musil, 1928:96 on the Rwala). Both the milk and the meat of the animals most commonly herded in the Near East - camels, sheep and goats - are used as food. While from the food point of view the camel is chiefly important for its milk, sheep and particularly goats are important, not only for their milk, but also for their meat. The Tuareg are a good example of how dependent the nomads of the Near East are on vegetable foods. It sometimes happens that for weeks, even months, on end they are forced to eat nothing but milk and then they complain of fatigue and stomach pains (Nicolaisen, 1963:209; cf. Monod, 1975:102). According to the data of Lhote (1955:213) the Tuareg of Ahaggar consume per head in a year 187 kilograms of millet, 15 kilograms of dates and an even greater quantity of wheat. The Arabs of the Sahara consume per head in a year 180 kilograms of wheat and 75 kilograms of dates. The Tuareg of Ayr consume per head half a kilogram of vegetable foods a day, or about 180 kilograms a year (Nicolaisen, 1963:213). Even today, despite the fact that Islam has held sway for many centuries, it has been recorded that nomads in Arabia and the Sahara do not preclude blood from their diet; this is true not only of the Tebu in the Sahara (Nicolaisen, 1963:312), but also of Bedouins in Arabia (Doughty, 1888, i:561; Finbert, 1938:39). Thus it may be assumed that blood was even more important in the past, for nowadays blood is only a supplementary element of the diet (see, for example, Nicolaisen, 1963: 235 on the Tuareg). Basically it is the Somali and the peoples akin to them who constitute the Northeast African sub-type of the Near Eastern type of nomadism. The species-composition of their herds is not very specific. The Somali also herd sheep, goats and camels, although they do not use the latter as riding animals (Zohrer, 1964-5:150). Donkeys serve as transport animals; there arc a very few horses, and cattle (zebu - Bos indicus) axe. more common in (he south, where pastoralism is combined with agriculture, than anywhere else (Lewis, 1955:329-30). Rainfall, particularly in North Somalia, varies from year to year and season to season. Consequently, although the annual cycle of two dry and two rainy seasons is regular, the routes of pastoral migrations are not stable (Lewis, 1955:331; Lewis, 1961:41). S7

Nomadism

as a distinct form of food-producing

economy

The food of the Somali consists first and foremost of milk and milk products, but meat is also an essential part of their diet, particularly in the dry season when there is insufficient milk. The slaughtering of livestock for meat is regular practice; some nomadic groups of the Somali eat the blood of camels and cattle. Vegetables are also a constant element in their diet (Swift, 1977:285). The Northeast African sub-type of the Near Eastern type of nomadism is in some ways similar to the nomadism of the East African type - the fact camels are not used as riding animals, the role of cattle and use of blood as food. This is particularly true of such groups as the Gabra who live in Kenya and Ethiopia to the north of the Samburu and the Turkana (Torry, 1976:269-71). But as a whole, both historically and from an economic point of view, the Northeast African sub-type is much closer to the Near Eastern type. 'Relations between Somaliland and Arabia are of great antiquity' (Lewis, 1955:45). Account must also be taken of the fact that in Somalia cattle has been most widely distributed in those areas where semi-nomadic pastoralism and herdsman husbandry is more common than pure forms of pastoral nomadism. In conclusion I should like to say something about one fairly large region the Sudano-Sahelian zone - which I cannot append to any one of the basic types of nomadism. If I were trying to make a more detailed typology then, possibly, I should have to append the region to one particular type, or even types, of nomadism (cf. Patai, 1978:16-20). However, since my aim is to single out only the basic types of nomadism, I cannot ignore the patent fact that this zone is an intermediate one and that in its basic parameters it is too lacking in uniformity. It is not surprising that Herskovits (1962:72-3), having singled out Eastern Sudan as a particular culture area in Africa, met with difficulties when trying to define it. Horowitz (1967) fundamentally doubted that this area in general could be considered a single culture area. At any rate the pastoral economy there (in which the nomadic economy is included) is not markedly homogeneous. As to the species-composition of herds, for example, the Kababish camelherders resemble nomads in North Africa (Asad, 1970) and the Baggara or Fulani cattle-herders resemble nomads in East Africa (Stenning, 1959; Cunnison, 1966). The Beja herd more cattle than camels (Paul, 1954:16). But where the character of pastoral migrations and the utilization of the products of pastoralism are concerned, there are essential differences between these groups and the nomads of East Africa, as well as of the Near East. For example, in contrast to East African nomads, the Humr Baggara migrate in regular seasonal cycles (Cunnison, 1966:13). The picture looks equally complex and contradictory in Sahel and Western Sudan. In Sahel zebu are herded alongside camels, goats and 58

The main types of pastoral

nomadism

sheep, and in the savanna cattle and sheep are becoming the most commonly herded animals and horses are beginning to acquire greater importance (Nicolaisen, 1963:28). South of the twelfth parallel there are very few camels (Awad, 1962:330). The pastoralism of the Tuareg in Niger is closer to that of the Fulani than it is to the pastoralism of the Tuareg in Ahaggar (Capot-Rey, 1962:303). Even where order of seasonal changes is concerned, Sahel is very different from North Africa and the instability of the pastoral migrations of nomads in the former is very marked (Bremaud and Pagot, 1962:318). As is the case in the Near East and East Africa, water is the most important economic priority in the migrations of nomads in the SudanoSahelian zone. Apart from general reasons for this which are linked with climatic conditions, there also are specific reasons, linked to the speciescomposition of herds, and which are characteristic only of the SudanoSahelian zone and of East Africa. In Sahel cattle cannot be pastured more than 15-20 kilometres from a watering place; if it is pastured further away it expends too much energy getting to the water (Bremaud and Pagot, 1962:320). The Humr Baggara in the Sudan move in such a way that water sources may always be found for their cattle (Cunnison, 1966:13). There are essential differences in dietary systems between pastoralists in the Sudano-Sahelian zone and pastoralists in East Africa. Thus the Fulani do not eat blood, but are clearly very fond of vegetable foods (Monod, 1975:143-4). As a result, in this book I am inclined to regard the Sudano-Sahelian zone as a marginal zone between the Near Eastern and East African types of nomadism. The Middle Eastern type. The nomadism of the Middle Eastern type in some respects is the intermediate type between the Near Eastern and Eurasian steppe types of nomadism. Considering its geographical location and the history of its formation, this is hardly surprising (see Chapter 2). The similarity between the nomadism of the Middle Eastern and Eurasian steppe types in certain areas of its distribution was noted back in the Middle Ages. At the beginning of the sixteenth century Babur wrote: 'just as Turks and Aimaqs roam in the open country of Khurasan and Samarqand, so in the Kabul area do the Hazara and Afghans' (The Babur-nama, 1922, vol. i:221; Zahir ad-Din Babur, 1958:165). Northwest Afghanistan may be looked upon as a marginal area between two types of nomadism (Ferdinand, 1969:129). Another such area is the territory to the north of Elburz in Northeast Iran. The economy of the Iranian Turkmen differed little from the economy of the Turkmen living in Russian Turkestan (Coon, 1976:220-1). Their migrations were separated only after the border between Russia and Iran was established in 1881 (Logasheva, 1976:35-6). The Turkmenian (South Middle Asian) sub-type itself is not internally 59

Nomadism

as a distinct form of food-producing

economy

uniform, although as a whole it can be looked upon as an intermediate one which in some respects is closer to the Middle Eastern type. However, since this sub-type was described in the section devoted to Eurasian steppe nomads, I shall not return to it here. At the same time Rashid al-Din recorded the similarity between Near Eastern and Middle Eastern nomadism. 'First of all one should know that in every zone of land there exists a separate [one from the other) population, [one] is sedentary, [another] nomadic. Particularly in an area [or country] where there are meadows and there is much grass, [in places] removed from the outlying parts of towns and from houses [settlements], often there are many nomads - we see this on the borders of Iran and in the domains of Arabs where there are waterless deserts with grass; such land suits camels because they eat much grass and use little water' (Rashid al-Din, 1952:73). From the Middle Ages groups of nomads who wintered in the Syrian desert and migrated for the summer to the high plateau of Asia Minor were well known. Somewhat later the Kurds also migrated in this way, after they had replaced the Turkic nomads in Eastern Anatolia in the sixteenth century (de Planhol, 1959:527-8). Finally, there are certain characteristics of the nomadism of small groups of the Kirghiz in Northeast Afghanistan which make it similar to the nomadism of High Inner Asia, the presence of yaks in the species-composition of the herds, for example (Nazif Mohib Shahrani, 1979). However, in contrast to the Sahara and Central Arabia and, even more so, to the Gobi and Kara-Kum, the deserts of the Middle East - Dasht-i Lut and Dasht-i Kavir - are without water and vegetation. Thus in the Middle East there are no real Bedouin, nor are there any pure nomads in the way that Mongols or Kazakhs were not so long ago. Al-Istakhri wrote that in Arabia and North Africa all the steppes, with the exception of separate and not very large stretches, are good pastureland and for this reason they are shared out between different nomadic tribes, while the Khurasan steppe is almost completely unpopulated; all the eye can see is road with postal stations along the sides (cited in Barthold, 1971:141). The Middle Eastern type of nomadism is probably the least uniform one. It is purely as a preliminary, an attempt to open the way for further research rather than to present conclusions, that I single out in the Middle Eastern type of nomadism the following sub-types: the western (mountainous), the main representatives of which are the semi-nomadic Kurds, Lurs and Bakhtiari; the northwestern, representatives of which are the Shahsevan and certain other Turkic groups in Iranian Azerbaijan ; the northern, representatives of which are Turkmen and other Turkic speaking peoples in Iran and Afghanistan, which, as has already been pointed out, can be looked upon as a marginal sub-type; the southern, based on meridional pastoral 15

15

In the past there were many nomads here (Petrushevsky, 1949.66). However, the majority of them are now sedentarized (Tapper, 1979a:101-3).

60

The main types of pastoral

nomadism

migrations of up to 500 kilometres, in which the summer pastures are in the mountains and the winter ones in the lowlands running down to the sea; the southeastern, basically semi-nomadic, the main representatives of which are the Baluch and Brahui who combine the herding of small stock and camels with the cultivation of date palms; and the eastern, mountainous, the main representatives of which are the nomads of Afghanistan. However, it should be pointed out that territorial borders and economic differences between the different sub-types are very fluid, and they are also not uniform. There are different forms of extensive pastoralism in the Middle East, which range from the purely nomadic to herdsman husbandry and which, as is also the case in the Eurasian steppes, are inter-reversible and interdependent (Trubetskoi, 1966:101; Coon, 1976:215). However, seden­ tary forms of animal husbandry with the stockpiling of fodder have not become widespread here (Barth, 1962:342). On the other hand, because of geographical conditions in the region, mountain variants of the cited forms have become particularly significant here. At present semi-nomadic pastoralism is far more common than pure pastoral nomadism in the Middle East. However, this is partly the result of changes in the modern period for in the Middle Ages purely nomadic forms of pastoralism were more widespread. In the species-composition of herds in the nomadism of the Middle Eastern type small stock is clearly predominant (Barth, 1962:344; Ferdinand, 1969:133), and this is from the point of view of quality as well as of quantity. Moreover, goats play a greater role in the species-composition of herds here than they do amongst the nomads of the Eurasian steppes, but their destructive influence on the landscape has, possibly, been exaggerated (Kolars, 1966). Amongst some nomads in the Middle East, the Baluch in South Persia, for example, goats play a greater role than sheep (Spooner, 1972:123; Spooner, 1973:8; Spooner, 1975:178-9). Amongst the Shah Nawazi there are two goats to every one sheep (Salzman, 1971:187), and in Makran the two species are to be found in approximately equal proportion (Pastner, 1971:174). The importance of camels and horses is limited; they are mainly used for transport, the horse first and foremost as a riding animal. In fact, the camel is of some essential significance only to the Baluch (Coon, 1976:192) and the Pathan (Spooner, 1975:174; Ferdinand, 1969:133). In Baluchistan the camel is used by semi-nomads for agricultural purposes (Swidler, 1973:32). It is interesting that although both breeds of camels, and also their hybrids, are to be found in the Middle East (Snesarev, 1921:87; Ferdinand, 1962:135; Fisher, 1966:300), today there are more dromedaries than Bactrians, which was not the case in the Middle Ages (Bulliet, 1975:161). As to the horse, in the literature it is recorded that only the Qashghai and Kurds 61

Nomadism

as a distinct form of food-producing

economy

specialize in breeding horses (Narody Serednei Azii, 1957:248; Patai, 1978:32). Before the beginning of the twentieth century, horse-breeding spread amongst the Bakhtiari (Trubetskoi, 1966:104). The donkey is very important as a transport animal (Hole, 1978:149). Cattle is not as a whole very important. In Khuzistan it is bred by the Bakhtiari and certain Arab tribes (Trubetskoi, 1966:12, 103). Because of environmental and geographical reasons almost nowhere in the Middle East is there any large ecological zone (large on the scale of Arabia, the Sahara or the Eurasian steppes) which is at the exclusive disposal of nomads. The territories occupied by nomads are ecological zones which in part have been artificially created, mainly in the Middle Ages as nomads ousted the agriculturalists from them and adapted them for the pasturing of livestock (see p. 104). Since agriculturalists and nomads were drawn into sharing certain ecological zones and niches in the Middle East the two have lived more intermingled here than they have in any other region. This shared utilization of zones and niches takes on different forms; however, they can be reduced to the two variants singled out by Barth which have been described above. Barth has also noted another characteristic of the extensive pastoralism (which includes pastoral nomadism) of the Middle East; this is the specific interconnection and interdependence between nomads themselves, because of their joint utilization on a rotational basis of the same ecological zones and niches. It is not surprising, therefore, that the routes of pastoral migrations of many nomads in the Middle East are very stable and regular (Spooner, 1972:124), for they depend not only on environmental and geographical conditions, but also on the complex system of both agriculturalists and other pastoralists utilizing different ecological niches on a rota basis. In the nomadism of the Middle Eastern type as a whole meridional, linear routes, or partially linear ones, are most common; and where the relief is mountainous, routes are vertical (or partially vertical) and of a clearly seasonal nature. Some groups of the Baluch are the only exception. In common with the majority of the other types of nomadism, the distance of pastoral migrations is very variable. Until quite recently it was as much as 1,000-1,500 kilometres amongst certain nomads in Southern and Eastern Afghanistan (Reisner, 1954:41-2; Ferdinand, 1969:133, 141). The dietary systems of the nomads of the Middle East are, in principle, similar to those of nomads in the Near East. Three elements - milk, vegetable produce and meat - form the basis of their diet. Amongst nomads in the Middle East the products of horses are less significant than they are to nomads in the Eurasian steppes, and the products of camels are less significant than they are to nomads in the Near East. Blood is not really 62

The main types of pastoral

nomadism

included in the diet. Literature about the Middle East would seem to suggest that at least some nomads here use less meat and more vegetables in their diet than do nomads in the Eurasian steppes; moreover, vegetables often constitute the main element of the diet (see, for example, Barth, 1962:345 on the nomads of Southwest Asia as a whole; Barth, 1964a:72 on the Basseri; Schurmann, 1962:261; Ferninand, 1969:136 on the nomads of Afghanistan; Salzman, 1971:188; Spooner, 1975:181; Swidler, 1973:30 on the Baluch). The East African type. Although the East African type of nomadism is not as homogeneous and monospecialized as the North Eurasian, it is nevertheless in many respects more uniform than either the Near Eastern or the Middle Eastern types. However, it does contain transitional forms. For example, the Rendille herd camels and the species-composition of their herds is similar to that of the Somali (Spencer, 1965:291; Spencer, 1973: Iff.). In his time Forde (1963:304) wrote of the almost complete absence of pure nomadism in this region and in many respects he was right. True, the Samburu (Spencer, 1965:22) and the Turkana (Gulliver, 1955:2) are pure nomads as also, apparently, is the majority of the Maasai. But many East African mobile pastoralists are also involved in agriculture, although the extent of this involvement is variable - for example, the Jie, the Nandi, the Dodos, the Pokot (Suk), the Mandary, the Kipsig, the Karimojong, the Iraqw, the Barabaig, and others (Huntingford, 1953:3, 4, 7, 77; Gulliver, 1955:51, 61; Deshler, 1965:158; Fukui, 1970: 101-4; R. and N. DysonHudson, 1969:78, 79; R. and N. Dyson-Hudson, 1970:114-17; Klima, 1970:13; Baker, 1975:189). Amongst East African pastoralists the species-composition of herds is very uniform. Cattle is without doubt the most important species but, despite the widespread stereotype, it does not monopolize the herds; small stock also always makes up part of the composition of herds (Spencer, 1965:3; Bohannan, 1966:129; Dahl and Hjort, 1976:220; Klima, 1970:104). The reasons why cattle predominates in the composition of herds in East Africa are not quite clear; presumably there are important ecological ones, apart from the cultural and historical. However, as yet they have not been properly researched. Some of the interesting ideas briefly suggested by Dahl and Hjort (1976:234) can be looked upon only as preliminaries. Another characteristic of the nomadism of the East African type is the absence of such transport and riding animals as the horse and camel. In West Africa the horse has reached the Guinea coast, but it is practically unknown in East Africa. South of Ethiopia horse-riding is unknown. Almost the same is true of the camel. Admittedly the Rendille herd camels, and the Turkana have begun to do the same relatively recently; but these peoples treat camels us they do cattle (Gulliver, 1955:39, 260-1) and cattle are not ridden in East 63

Nomadism

as a distinct form of food-producing

economy

Africa (Forde, 1963:401). Meanwhile, in the Sudan the Baggara use bulls as transport animals (Cunnison, 1966:31). Donkeys are sometimes used as transport animals, although only to a very limited extent. Thus the Barabaig and the Karimojong use donkeys as beasts of burden, but do not ride them (Klima, 1970:100; R. and N. Dyson-Hudson, 1970:100). It may be said, with certain reservations, that East African pastoralists play the dominant role in their ecological zone. As has already been pointed out, the majority of pastoralists here are also involved in agriculture, but basically in separate enclaves. According to Baxter (1975:212): '. . . at any rate since the establishment of colonial rule, Kenyan pastoralists move within their own demarcated tribal territory and do not follow a seasonal cycle through the territories of other sedentary peoples.' The same state of affairs evidently existed before colonial rule (see, for example, Alpers and Ehret, 1975:492 on the Maasai and the Sonjo). As is not the case with regard to the Middle East and North Africa there is no need to discuss any joint or rotational utilization of the ecological zone or niches in East Africa. The character of pastoral migrations is determined first and foremost, and more than anything else, by the geographical and climatic particularities of East Africa, especially the irregularity of precipitation. Consequently, the routes of pastoral migrations are irregular, unstable, spread out and nonlinear (see, for example, Gulliver, 1955:38; cf. Gulliver, 1975; Spencer, 1965:7; R. and N. Dyson-Hudson, 1969:89 on the Jie, Turkana, the Samburu and the Karimojong). The particularities of the species-composition of the herds of the nomads of East Africa must also be taken into account when the character of their pastoral migrations is examined. Unlike camels, cattle can go without water for only a limited period; amongst the Samburu, in the most extreme circumstances, the cattle are watered every third day (Spencer, 1965:6). Cattle are usually pastured separately from small stock, but sheep and goats are often pastured together (Dahl and Hjort, 1976:250). The distance covered by pastoral migrations is really not very great. Milk is the basic and most important element in the diet of East African pastoralists. Of pastoral products meat takes second place (at least from a calorific point of view), and blood third place. However, meat plays a less important role in the diet of East African nomads than it does in the diets of nomads in many other regions. Blood contains the iron which is lacking in milk; but, most importantly, blood is a source of food in the dry season when supplies of milk are insufficient. However, the calorific value of milk and meat is much higher than that of blood (Deshler, 1965:162; Dahl and Hjort, 1976:173-5). It would appear that pastoralists of the East African type use less vegetable foods than nomads in the Near and Middle East, but for practically every pastoral group in this region vegetables are an indispen64

The main types of pastoral

nomadism

sable part of the diet (Schneider, 1964:68). Literature on the subject, particularly the older literature and partly also the contemporary (see, for example, Baxter, 1975:214), can sometimes give the impression that their diet is made up solely of the products of pastoralism. However, there are indications that pastoralists in East Africa use considerably more vegetable foods than they are prepared to admit. An observation made by Bohannan (1966:132) seems noteworthy: 'Actual studies by food economists, dieticians and geographers, however, have recently indicated that some starchy staple diet enters the diet of almost all African herdsmen.' This is, indeed, the case. For example, Deshler (1965:165-6) clearly states that cereals constitute a most important part of the diet of the Dodos (Dodoth). Regular use of vegetable foods has also been noted amongst the Karimojong (R. and N. Dyson-Hudson, 1969:79) and the Barabaig (Klima, 1970:13). With the Maasai the case is less clear. Forde (1963:297) has written that all the Maasai, excepting the warriors, eat cereals and root-plants. Meanwhile Jacobs (1975:407-9) insists that amongst the Maasai there exist strict prohibitions of any non-pastoral food; however, he also notes that in this respect the Maasai are an exception amongst East African pastoralists. There does exist the opinion that this prohibition on all foods but the products of pastoralism appeared only amongst certain southern groups of the Maasai in the seventeenth and eighteenth centuries (Alpers and Ehret, 1975:491-2). One more fact deserves attention. This is that between the different age and sex groups of pastoralists of the East African type there exist significant variations in use of vegetable food; for example, the women use more vegetable food than the men do (R. and N. Dyson-Hudson, 1970:100). The High Inner Asian type. The principal representatives of the High Inner Asian type are the nomads of Tibet (Kussmaul, 1962; Ekvall, 1968). The Kirghiz of the Pamirs constitute one notable sub-type. The environments of the Pamirs and Tibet have much in common (Agakhaniants, 1965:16ff.). In common with almost all other types of nomadism, neither from the geographical nor from the economic and cultural point of view, is this type either fully isolated or fully homogeneous. In the northeast, where the Tibetan plateau is bordered by the plains of Northwest China and the Mongolian steppes, nomadic herds include camels, horses, sheep and goats, and in this respect they are similar to the nomadism of the Eurasian steppe type. Amdo (Tsing hai) is also a marginal area from the ethnic point of view, for Mongols, Salars, Kazakhs and other nomads of the Eurasian steppes live there alongside the Tibetans (Hermanns, 1949:29-34). But in other areas the specificity of the nomadism of the High Inner Asian type can be traced very clearly. 65

Nomadism

as a distinct form of food-producing

economy

Again, as in the case almost everywhere else, in Tibet semi-nomads are to be found alongside pure nomads (Downs, 1964:1115-19; Downs and Ekvall, 1965: 180-2); both principal variants of semi-nomadism are present, as also are groups involved in herdsman husbandry in its mountain yaylag variant. The specificity of the nomadism of this type is contained in the distinctive composition of herds, in which the most important species are the yak and a particular mountain breed of sheep; in most areas these are to be found in a proportion of 1:4-8 (Ekvall, 1968:11-12). The yak is also the most important transport animal and can carry packs until it is thirty years old (Potanin, 1950:308), although in Western Tibet sheep are even used as pack animals (Hermanns, 1949:84). Apart from in Tibet the domestic yak is also to be found in Mongolia, Manchuria, the Trans-Baikal area, the Sayan and Altai mountains and in Tien-Shan. But only in the Pamirs is the yak as important as it is in Tibet. Amongst the Kirghiz of the Eastern Pamirs in the nineties of the last century there were 7-8 thousand yaks to 40-50 thousand small stock (Shibaeva, 1973:104). In Tibet cattle and horses are much less important. There are relatively few horses and the Tibetans do not eat the meat or drink the milk of their horses, nor do they use them to carry packs. On the other hand, horses are prestigious animals and important for raiding and warring. Cattle are ill-suited to high mountainous areas and are looked upon as the animal of poor agriculturalists. Hybrids, however, especially of the yak and cow, play an important role in the economy as they can survive in high mountains and also at a relatively low altitude (Downs and Ekvall, 1965:176-9). The composition of the herds of the nomads of Tibet in some ways is essentially different from that of the semi-nomads here. The semi-nomads herd fewer yaks and sheep, but more cattle and sometimes specialize in the herding of the latter. The nomads of High Inner Asia occupy the dominant position in their ecological zone which, vertically, is very clearly separated from the zone where there is agriculture. This factor explains why there is almost no competitiveness over the utilization of the zones. 'Altitude is the prime factor, separating grazing land from farming land' (Ekvall, 1968:5). Where the pastoralists roam agriculture is impossible. The lower the altitude and the better the conditions for agriculture, the worse the yak feels; at lower than 10,000 feet it is incapable of breeding. The Kirghiz of the Pamirs also herd yaks only in the most mountainous areas, but those groups of the Kirghiz who descend into valleys, if only for the winter, are unable to breed the yak (Lus, 1930:155). However, the borders separating nomads from agriculturalists in Tibet 66

The main types of pastoral

nomadism

are variable. In Northeast Tibet 3,000 metres is about the highest limit for agriculture (Potanin, 1950:257), but in Southern Tibet the monsoons which push through the Himalayas from India make possible the cultivation of barley at a height of at least 4,200-4,600 metres, and of wheat at 3,900 metres (Kovalevsky, 1938:486-7). Consequently, the pastoralism of Tibet in the main is concentrated in the northwest, northern and northeast parts of the country (Kychanov and Savitsky, 1975:122). The pastoral migrations of the nomads of Tibet are regular; they occur in specific seasonal cycles and in a vertical direction, and availability of fodder according to the time of year is the most important economic priority. However, the routes of pastoral migrations here are less stable than those of the Eurasian steppe type of nomadism. There are several explanations for this. Owing to the mountainous landscape in Tibet, the pasturelands are not in one continuous territory; the high altitude and extreme continental climate subject the country to frequent and irregular microvariations, according to which the suitability of the pastures changes; yaks cannot independently get at fodder in snow covered pastures, thus the nomads have to look for pastures on steep places from which the wind has blown the snow, or in rocky spots where there is no snow (Baskin, 1976:115). The dietary systems of the nomads of High Inner Asia also demonstrate very clearly the specificity of this type of pastoralism. It is in Tibet that the most composite utilization of the three traditional products of a pastoral economy- milk, meat and blood - is to be observed. The climatic conditions in mountain areas increase man's need for food high in protein. Consequently, out of all the different types of nomadism, with the exception of the North Eurasian type, the consumption of meat is highest in the High Inner Asian type. At the same time the consumption of blood is here higher than it is in any other type of nomadism, with the possible exception of the East African type. But I do not have precise comparative data on this to hand. The meat consumed has consisted basically of mutton and yak-meat, and the milk consumed has been that of the yak. According to Ekvall (1968:59), '. . . the foodstuffs derived from these three sources (milk, meat, blood) are, in themselves, and without additives of fruit, cereals, or vegetables, an amply nutritious diet.' Nevertheless, vegetable foods have been a constant element in the diet of the nomads in Tibet. In the Eastern Pamirs the sheep has been the principal animal for supplying meat (Shibaeva, 1973:108). In contrast to the Tibetans, the Kirghiz are not accustomed to drinking fresh milk. In conclusion I should like to say something about two regions where pastoralism manifesting a series of specific characteristics exists. However, it is not my intention to examine this pastoralism in detail for it cannot be categorized as a pastoral nomadic type. The idea that there does exist a particular Southern Asian (Indian) type of 67

Nomadism

as a distinct form of food-producing

economy

nomadism raises serious doubts. There are evidently no pure nomads in India today, just as there are no sizeable groups of semi-nomads. It is not rare for certain groups, in Rajasthan and Gujarat, to be referred to as semi-nomads (Bose, 1975). But these groups do not drive their livestock every year to pastures far from their place of settlement, and only between 3.5-37.4% of the households participate in such migrations; moreover, not all members of the households involved move with the livestock (Bose, 1975:5-8). At best the pastoralism here can be regarded only as herdsman husbandry; however, when the species-composition of herds and certain other characteristics are taken into consideration, it certainly can be looked upon as a sub-type of the pastoralism of the Middle Eastern type. The Indian pastoralists living further south who inhabit the forest and mountainous regions of Central India are primarily cattle-breeders. They share part of one ecological zone with agriculturalists, but have no autonomous economic system of their own. Essentially their economy is only an appendage of the agricultural one (see, for example Sontheimer, 1975). On the subject of 'pastoral castes' in India Sopher (1975:195) writes: 'Two qualifying statements have always been understood to apply to this categorization: (1) many members, usually a large majority, of such so-called pastoral castes are primarily cultivators or follow some other nonpastoral occupation as their chief livelihood; (2) people of castes or other communities whose traditional function is not pastoralism as such may pursue pastoral occupations, becoming predominant in this activity in some areas.' Judging from data provided by Mahapatra (1975:210-11) such pastoral castes as the Ahir, Goala, Gonda, Yadava (Jadav), Gavli (Gowli), Ghosi, Kolari, Kurula, Dhangar, Mirdha, Kurmar, Kurumba, Mainiyani and others who, as a rule, are involved in agriculture in a supplementary capacity and even such tribes as the Toda should be included in the list of sedentary and semi-sedentary pastoralists. The Gaddi, Bharavad, Rabari (Rehbari), part of the Gavli (Gowli) and Dhangar, the Golla, Gauda and others are involved in herdsman husbandry (in its yaylag variant in the foothills of the Himalayas) and have permanent dwelling places in villages. Almost the same doubts are raised by the idea that there exists an Andes type of nomadism. There are known to be groups of llama breeders in the Andes today who have nothing to do with agriculture (Webster, 1973:116,117). However, the majority of groups herding llamas and alpacas in the Andes combine pastoralism with agriculture. Their pastoralism is related to the mountain variant of herdsman husbandry (Webster, 1973; Browman, 1974; cf. Murra, 1965:188). In the pre-colonial period in the Andes pastoralism was closely linked with agriculture on account of many different circumstances. Amongst the 68

Problems

of balance and non-autarky

of the pastoral

economy

latter were absence of the consumption of milk and milk products, the slow rate at which the llama and alpaca reproduce, and the fact that the majority of families had herds of less than the 150-200 head which was the required minimum for maintaining a purely pastoral economy (Murra, 1965:188; Browman, 1974:190). Problems of balance and the non-autarky of the pastoral economy The peculiarities of pastoral nomadism have been examined and the question of its specificity may once again be approached. Conditionally, all forms of pastoralism may be regarded as different methods of economic adaptation, the parameters of which are determined, in the final analysis, by ecology and level of technological development. For all this, pastoral nomadism is something which is very specialized, both from the economic and, partly, the cultural points of view. Pastoral nomadism can be looked upon as an answer to conditions dictated by environment and, moreover, it is a successful answer because out of all forms of the traditional food-producing economy it was pastoral nomadism which was able to master and exploit the potential resources of vast ecological zones. The emergence of pastoral nomadism was an important step in the spreading of a food-producing economy in the arid, semi-arid and tundra zones of the oikoumene. While social and political factors are temporarily left aside, it can be said that pastoral nomadism, and sometimes semi-nomadic pastoralism as well, developed and functioned in the first instance in those regions where they had economic advantages over all other local kinds and forms of economic activity. Furthermore, in most regions in the zones examined in this book, this advantage lasted right up to the twentieth century; in some places it is even preserved in the present day. In antiquity Strabo (Strabo, VII.4.6) recorded that even in the Crimea, one of the most fertile areas in the Eurasian steppes, the harvest only yielded thirty-fold, while in Mesopotamia it yielded three-hundred-fold. Even today extensive pastoralism yields more in many areas than agriculture does (Shulzhenko, 1954:217ff.). In Middle Asia herdsman husbandry in which herds are pastured all year round is a less labour-consuming and more profitable form of economy, 1 \ times more lucrative, than cotton-growing (Fedorovich, 1973:218). However, the limited possibilities for innovation in the economic system < if pastoralism are also very evident. If, from the point of view of technology, many forms of agriculture demonstrate they are capable of progressive diachronic technological changes, whatever the reasons were for these changes, in pastoral nomadism, on the other hand, once its formation is complete, the simple reproduction of highly specialized forms of the same type prevails. This does not mean that nomadism as a whole is a highly 69

Nomadism

as a distinct form of food-producing

economy

specialized blind alley, or that each of its variants and sub-variants is such. It is just that in the wider sense nomadism cannot be fully identified with pastoral economy, although the latter forms the basis of nomadism. The ecological foundations of pastoral nomadism are such that they leave little scope for the development of a complex economy in zones where pastoral nomadism is widespread; only in certain cases do they permit specific variations in pastoralism (for example, variations in speciescomposition of herds). On the one hand, it is this specialization which has enabled nomadism to occupy the dominant position in many ecological zones and to utilize these zones economically over a long period. On the other hand, this specialization limits opportunities for economic growth which is to the cost of productivity increase in the pastoral economy of nomadic societies, even when these societies are socially stratified and politically centralized. With rare and incomplete exceptions, specialization in a primitive and traditional economy sharply differs from specialization in the market economy of an industrial society, partly for the reason that the former occurs within the framework of a subsistence-oriented economy. However, a specialized pastoral economy, in contrast to the economy of many comparable forms of agriculture, itself cannot provide even all the immediate requirements of nomads. Nomadism is practically inseparable not only from supplementary forms of economic activity, but also from such social and political activity which numbers amongst its aims the overcoming of economic one-sidedness. Of course, in the different variants of nomadism and even in different nomadic societies, specialization has manifested itself in different ways and with different degrees of intensity. But as a general rule the well-known aphorism of Lattimore (1967:522): 'It is the poor nomad who is the pure nomad' is scarcely true. It so happens that impoverished nomads have been forced more often than other nomads to seek supplementary sources of subsistence and in the process not infrequently have become sedentary. However, in the final analysis, the notions of 'poverty' and 'wealth' are relative, to a certain extent subjective, and they vary from age to age and society to society, depending on different historical and socioeconomic factors, cultural values and traditions, etc. Consequently, much more important and objective are other economic problems which specialization imposes on pastoral nomadism. The most important problem noted by Barth and many other scholars is that of the balance between availability of natural resources (fodder and water), number of livestock and population-size (Barth, 1964:123ff.; Sweet, 1965:137; Paine, 1971:161; Swidler, 1973:23-4). Barth (1959-60:8) maintains that: 'Unless techniques for the storage of fodder are developed, absolute population-size is limited by the carrying 70

Problems

of balance and non-autarky

of the pastoral

economy

capacity of the pastures in the least productive period of the year.' Whether or not the least productivity of pastures should be taken into account when the dependence between population-size and natural resources is consi­ dered remains a debatable point. Domestic animals, particularly camels, are able to store up energy acquired in a period of plentiful and fully nutritious feeding, and nomads (at least some nomads) know how to regulate size of herds according to the season and how to lay in stores of the products of pastoral economy. Barth (1959-60:9), writing of the nomads of South Persia, cites examples of how their system of pastoral migrations is based on utilization of pastures when they are at their most productive. The fact is that between numbers of nomads and fodder resources they have at their disposal there exists a specific dependence, and the intermediary factor in this dependence is number of livestock. Nomads themselves empirically understand this dependence very well. According to the Mon­ gol proverb: 'Without grass there is no livestock, without livestock no food.' As Bates and Lee (1977:828) have rightly pointed out, the problem of balance takes on different forms according to the extent of involvement of nomads in market economy. In effect, the problem does look very different amongst the Karimojong or the Turkana to the way it does amongst the Somali today (Swift, 1977), or amongst the nomads of Makran, for the latter sell 25% of the annual increase in the number of their livestock and in exchange buy agricultural products and handicraft articles (Pastner, 1971:177). However, this close dependence on market trading is only characteristic of nomads in the modern period, and even in the present day not by any means is it characteristic of all nomads. Pastoral nomadism is doomed to stagnation because its economy is extensive and allows no permanent solution to the problem of balance at the expense of intensification of production. According to the calculations of a number of scholars, the number of livestock per head amongst the Hsiung-nu, who in ancient times occupied the territory which now constitutes Mongolia, corresponds almost exactly with the number which was found amongst the Mongols in Autonomous Mongolia in 1918: the figures were 19 and 17.8 head respectively (Egami, 1956, non vidi; Taskin, I968a:41ff.). It is remarkable that although the population of Mongolia in the sixties was double what it was in 1918, the number of pastoralists in the steppe itself remained practically unchanged. The population increase took place in the towns (Markovska, 1973:290). A relatively small population increase, smaller than the increase amongst agriculturalists, is characteristic of many African nomads (Dahl and Hjort, 1979:7-8). Some of them have even consciously practised birth-control, particularly in periods of drought (Swift, 1977:282). The Rendille employ different methods of birth-control, Including infanticide; moreover, since the balance is constantly being upset, 71

Nomadism

as a distinct form of food-producing

economy

a steady flow of the Rendille population to the Samburu is to be observed (Douglas, 1966:263, 270). In order to increase the productivity of pastures, a whole series of intricate and complex measures must be taken, and in the conditions in which pastoral nomadism existed until not so long ago such measures were practically impossible. However, even today certain measures which aim to facilitate the maintaining of the traditional extensive pastoral economy with contemporary technical methods often do not achieve their aim, and sometimes their consequences are unexpected and undesirable. Thus in many countries the drilling of boreholes irrespective of grazing leads to exhaustion of pastures through overgrazing (Glantz, 1976; Norton, 1976:260; Le Houerou, 1976:270; Le Houerou, 1977:28). Nomads are sometimes reproached for not laying-in fodder for their livestock in winter. But in the Eurasian steppes herbage takes several years to grow again after it has been cut, during which time the hay fields often become overgrown with shrubs and lose their pasture value (Kalinina, 1974:15). There is yet another reason - lack of the necessary workforce which prevents the large-scale laying-in of hay. 'A herd of 100 horses can be tended all year round by one person, but to feed the same number of animals in winter on hay 15,000 poods [1 pood = 16.38 kilograms] of hay would be required' (Dmitriev, 1903:100). In a nomadic economy only a temporary balance is possible. The only way in which a balance can be maintained is by the constant manipulation and adjustment of all three of the above mentioned variables. Only by constantly altering them can they be kept in an optimal relationship to each other. Even microfluctuations play an important role in maintaining such an equilibrium. Thus, in the USSR, the crop capacity of wild plants in semi-deserts fluctuates between 1:5.4 in different years, and of different grass-crops it even fluctuates as much as 1:40; in deserts it fluctuates between 1:5, and in an annual grass-crop as much as 1:60. All this has an adverse effect on pastoralism, even in the present day. There were 5.1 million head of sheep and goats in Turkmenia in 1959-60, in 1962 the number had dropped to 4.2 million (Fedorovich, 1973:216). The situation is complicated still further by the fact that these variables are not engendered simultaneously or by factors operating synchronically. Each of them in turn is determined by many factors, some constant and some temporal in character, some regular and some irregular ones. Barth (1962:350) writes of the nomads of Southwest Asia: 'We know that population-movements, periodic famines and epidemics, as well as unpaired fertility together through the centuries have tended to keep the total population fluctuating around a roughly stable average.' This is true. But what is also true is that fluctuations in population-size have not always coincided with corresponding fluctuations in numbers of livestock (see, for 72

Problems

of balance and non-autarky

of the pastoral

economy

example, Khazanov, 1979; Khazanov, 1980a on twelfth century Mongolia). Characteristic of the nomads of North Eurasia at the end of the nineteenth century and the beginning of the twentieth was a specific type of demographic structure and population growth, in which there was an extremely high birth-rate, a very high infant mortality rate and a low, but steady, population increase. Amongst the Nentsy towards the beginning of the twentieth century (and possibly even earlier), this led to a situation in which the balance between resources in the territory they utilized and population-size was upset. As a result pastures were exhausted and the Nentsy had to resort to buying great quantities of foodstuffs. According to the calculations of Krupnik (1976:80ff.), in the twenties of the present century purchased food constituted 45-60% of the energy value in the food balance of the Nentsy. Of course, high birth- and death-rates are also characteristic of many traditional agricultural societies. However, neither so quickly, nor invariably have they led to the consequences they did amongst the Nentsy. Climatic pulsations of varying duration, the presence of which is now universally recognized in climatology, have directly affected natural resources utilized by livestock and helped to upset the balance in question. This is not all. It is now known that at least some populations of domestic animals are subject to specific biological pulsations which, possibly, are somehow connected to climatic pulsations. For example, the number of deer amongst the nomads of North Eurasia undergoes very clear fluctuations, with recurrent sharp falls of up to 50% every 10-15 years, due to large-scale outbreaks of epizootic disease in the Eurasian North or the freezing over of pastures in Chukotka (Krupnik, 1976:113). The Kazakhs believe there are 'difficult years' in the twelve-year animal chronological cycle; such years are those of the mouse, hare and cock. But the most dangerous is thought to be every third year of the hare, which is accompanied by severe jute (Narody Srednei Azii, 1963:359). Finally, the simplest and, therefore, the most well-known reason for a temporary upset of the balance is when huge numbers of livestock perish as a result of natural disasters or epizootic diseases. In 72 B.C. there was famine amongst the Hsiung-nu on account of heavy snowfalls and the successful raids of hostile peoples. 'Three people out of every ten died, five animals out of every ten fell' (Taskin, 1973:28). Famine recurred in 68 B . C . : 'In the same year in the lands of the Hsiung-nu there was famine, because of this six or seven people out of every ten died, six or seven animals out of every ten fell' (Taskin, 1973:29). In Kazakhstan in the nineteenth century jute was a serious problem and in some years caused between half and three-quarters of the overall number of livestock to perish; large-scale losses occurred every 6-11 years and local ones almost annually (Tolybekov, 1959:54-6; cf. Ishchenko et al., 1928:69 7.1

Nomadism

as a distinct form of food-producing

economy

a steady flow of the Rendille population to the Samburu is to be observed (Douglas, 1966:263, 270). In order to increase the productivity of pastures, a whole series of intricate and complex measures must be taken, and in the conditions in which pastoral nomadism existed until not so long ago such measures were practically impossible. However, even today certain measures which aim to facilitate the maintaining of the traditional extensive pastoral economy with contemporary technical methods often do not achieve their aim, and sometimes their consequences are unexpected and undesirable. Thus in many countries the drilling of boreholes irrespective of grazing leads to exhaustion of pastures through overgrazing (Glantz, 1976; Norton, 1976:260; Le Houerou, 1976:270; Le Houerou, 1977:28). Nomads are sometimes reproached for not laying-in fodder for their livestock in winter. But in the Eurasian steppes herbage takes several years to grow again after it has been cut, during which time the hay fields often become overgrown with shrubs and lose their pasture value (Kalinina, 1974:15). There is yet another reason - lack of the necessary workforce which prevents the large-scale laying-in of hay. 'A herd of 100 horses can be tended all year round by one person, but to feed the same number of animals in winter on hay 15,000 poods [1 pood = 16.38 kilograms] of hay would be required' (Dmitriev, 1903:100). In a nomadic economy only a temporary balance is possible. The only way in which a balance can be maintained is by the constant manipulation and adjustment of all three of the above mentioned variables. Only by constantly altering them can they be kept in an optimal relationship to each other. Even microfluctuations play an important role in maintaining such an equilibrium. Thus, in the USSR, the crop capacity of wild plants in semi-deserts fluctuates between 1:5.4 in different years, and of different grass-crops it even fluctuates as much as 1:40; in deserts it fluctuates between 1:5, and in an annual grass-crop as much as 1:60. All this has an adverse effect on pastoralism, even in the present day. There were 5.1 million head of sheep and goats in Turkmenia in 1959-60, in 1962 the number had dropped to 4.2 million (Fedorovich, 1973:216). The situation is complicated still further by the fact that these variables are not engendered simultaneously or by factors operating synchronically. Each of them in turn is determined by many factors, some constant and some temporal in character, some regular and some irregular ones. Barth (1962:350) writes of the nomads of Southwest Asia: 'We know that population-movements, periodic famines and epidemics, as well as unpaired fertility together through the centuries have tended to keep the total population fluctuating around a roughly stable average.' This is true. But what is also true is that fluctuations in population-size have not always coincided with corresponding fluctuations in numbers of livestock (see, for 72

Problems

of balance and non-autarky

of the pastoral

economy

example, Khazanov, 1979; Khazanov, 1980a on twelfth century Mongolia). Characteristic of the nomads of North Eurasia at the end of the nineteenth century and the beginning of the twentieth was a specific type of demographic structure and population growth, in which there was an extremely high birth-rate, a very high infant mortality rate and a low, but steady, population increase. Amongst the Nentsy towards the beginning of the twentieth century (and possibly even earlier), this led to a situation in which the balance between resources in the territory they utilized and population-size was upset. As a result pastures were exhausted and the Nentsy had to resort to buying great quantities of foodstuffs. According to the calculations of Krupnik (1976:80ff.), in the twenties of the present century purchased food constituted 45-60% of the energy value in the food balance of the Nentsy. Of course, high birth- and death-rates are also characteristic of many traditional agricultural societies. However, neither so quickly, nor invariably have they led to the consequences they did amongst the Nentsy. Climatic pulsations of varying duration, the presence of which is now universally recognized in climatology, have directly affected natural resources utilized by livestock and helped to upset the balance in question. This is not all. It is now known that at least some populations of domestic animals are subject to specific biological pulsations which, possibly, are somehow connected to climatic pulsations. For example, the number of deer amongst the nomads of North Eurasia undergoes very clear fluctuations, with recurrent sharp falls of up to 50% every 10-15 years, due to large-scale outbreaks of epizootic disease in the Eurasian North or the freezing over of pastures in Chukotka (Krupnik, 1976:113). The Kazakhs believe there are 'difficult years' in the twelve-year animal chronological cycle; such years are those of the mouse, hare and cock. But the most dangerous is thought to be every third year of the hare, which is accompanied by severe jute (Narody Srednei Azii, 1963:359). Finally, the simplest and, therefore, the most well-known reason for a temporary upset of the balance is when huge numbers of livestock perish as a result of natural disasters or epizootic diseases. In 72 B.C. there was famine amongst the Hsiung-nu on account of heavy snowfalls and the successful raids of hostile peoples. 'Three people out of every ten died, five animals out of every ten fell' (Taskin, 1973:28). Famine recurred in 68 B.C: 'In the same year in the lands of the Hsiung-nu there was famine, because of this six or seven people out of every ten died, six or seven animals out of every ten fell' (Taskin, 1973:29). In Kazakhstan in the nineteenth century jute was a serious problem and in some years caused between half and three-quarters of the overall number of livestock to perish; large-scale losses occurred every 6-11 years and local ones almost annually (Tolybckov, 1959:54-6; cf. Ishchenko et ai, 1928:69 73

Nomadism

as a distinct form of food-producing

economy

who cite more moderate figures). In Mongolia at the beginning of the twentieth century up to 50% of the livestock perished from epizootic diseases (Viatkina, 1960:151). Half the livestock of the nomads of South Persia could perish in one season (Barth, 1964a:71). Between half and one-third of the overall number of the reindeer of the Chukchi could perish in one summer (Bogoras, 1904:80; Paine, 1971:162). Gulliver (1955:103, 122) records the stability of the Jie population and explains the cost at which this stability is achieved: 'A serious food-shortage occurs about once a decade, and less frequently there have been severe famines, both of which cause varying losses of life and also a weakening of general physical condition and resistance to disease. Occasionally there bad years coincide with stock-epidemics' (cf. Peters, 1960:37 on the stability of the nomadic population of Cyrenaica). Finally, mention must be made of the recent drought in Sahel and its catastrophic consequences - the perishing of one hundred thousand people and no less than 25% of the overall number of livestock - especially since it was already the third such drought to have taken place during the course of this century (The Politics of Natural Disaster, 1976; Glantz, 1977:2-3; Grove, 1978). Even in Tibet where, according to Ekvall (1968:19), there is an abundance of pasture, numbers of livestock and increases in the overall number of stock conform to the same tendencies which exist in other regions where there is pastoral nomadism. There are sayings amongst many nomads which reflect the instability of their economy. When economies are compared it becomes clear that everywhere nomads are faced with the same problems. Thus a Somali saying: 'Abundance and scarcity are never far apart; the rich and the poor frequent the same house' (Lewis, 1961:37) may be compared with a Kazakh one: 'Livestock belongs to any snowstorm and powerful enemy' (Tolybekov, 1959:218) and with a Mongol one: 'One jute suffices the rich man, and one arrow the hero' (Zhagvaral, 1974:113). If homoeostasis is taken to mean the striving of a system towards constancy in all of its functional states and connections, then a nomadic economy can only conditionally be called homoeostatic, and that only with regard to tendencies of specific cycles in its inner development. The equilibrium of a nomadic economy is not static; it is dynamic and it is maintained in a specific way. It could be said that the balance also includes the conformity of available workforce to quantity of livestock (Paine, 1972:78), but this problem is usually solved with the help of corresponding social mechanisms (see Chapter 3). The notion that nomads tend to strive towards ecological equilibrium and the artificial preservation of economic-demographic stability in the conditions of a permanent environment and of available resources which 74

Problems

of balance and non-autarky

of the pastoral

economy

change little with time, bears little relation to reality; this is because, as may be observed at least in the cited examples, such a situation is simply impossible. On the contrary, more often than not what is characteristic of nomads is heightened tension in all economic and demographic processes, and in favourable periods a high birth-rate and striving towards increased production; it is only in these circumstances that nomads themselves and their economy are able to stand up to natural disasters. As a result the alternation of favourable and unfavourable periods produces its own kind of levelling, and regulation of the homoeostasis of the system itself takes on the form of cyclical variations. The attitude of nomads to pastures is linked to the distinctive way of maintaining the equilibrium. The view that nomads are not concerned about preserving pastures (Spooner, 1973:16, 24) as a general rule scarcely corresponds to reality. There are well-known examples where exactly the opposite is true. Thus, reindeer-herders have sought to ease the load on the pasture resources they utilize (Krupnik, 1976:93). The Bedouin, up until the last possible moment, hold back from utilizing their summer pastures (Marx, 1978:49; cf. Digard, 1973:47-8 on the Bakhtiari). Amongst the Somali the traditional leaders in the north of their country had the power to prohibit the utilization of specific pastures in order that the grasses in them should grow again (Swift, 1977:284). Some nomads really do not seem to be concerned about pastures and where this state of affairs is not linked to social and political factors as, for example, amongst the Karimojong (R. and N. Dyson-Hudson, 1969:79), or in North Kenya in the sixties of the present century (Dahl and Hjort, 1979:8), or amongst the Kazakhs of the Bukei (Inner) Horde in the nineteenth century (Shakhmatov, 1964:64-5), it may be explained by the fact that the nomadic economy in question was in a favourable cycle of its development. In the Sahel in years when the quantity of precipitation is favourable, the number of animals and people increases so much that it begins to exceed the carrying capacity of marginal rangelands in less favourable years (Katz and Glantz, 1977:96-7). An unfavourable period, which inevitably follows a favourable one, automatically eases the load on pastures and can lead to their partial or complete renewal. In Mongolia, in strips of pasture where livestock has not pastured for some five years, the crop-capacity of the grasslands reaches 10 centners [1 centner = 100 kg] a hectare, while that of grasslands which have been pastured is only 3.2 centners (Zhebrak, 1933:6). Swift (1977:285) notes of the Somali: 16

17

'I'he cited data on African pastoralists proves that a high birth-rate is not characteristic of all nomads. However, it is not inconceivable that this data relates to unfavourable periods. At nny rale there is too little available data for any precise conclusions or comparisons to be made. Moreover, birth-rate should not be confused with population-growth. " Such a model in certain respects is similar to that suggested by Krupnik (1976) for the traditional economy of the peoples of the tundra zone (including reindeer-herders), which is imiiihlc for extreme variability of ecological conditions.

75

Nomadism

as a distinct form of food-producing

economy

'Periodic droughts and disease epizootics decimated the herds regularly and generally kept them in a long-term dynamic equilibrium with the vegetation's ability to regenerate itself.' When the distinctive form of a pastoral economy and the particularities of its self-regulation are ignored, wrong conclusions are sometimes reached about the attitude of nomads to livestock reproduction on an expanded scale. Some scholars believe that maximization of numbers of livestock is the most important trend in the economic strategy of nomads, almost an end in itself. For example, Paine (1971:168-9) writes that herd expansion is 'a basic pastoral value' and 'is often practised whether or not there is opportunity for corresponding expansion of pastures' (cf. pp. 72-3 on reindeer-herders). However, nomads and anthropologists studying them would do better not to think along such lines; in practice long-term maximization of numbers of livestock is only possible if it is linked with specific factors outside nomadic society. Of course, what we are talking about here are objective economic and demographic processes in nomadic society as a whole, and not about the aims, actions and even the results of the actions of separate individuals, families and groups within nomadic society. As Imperato (1976:286) writes of African pastoralists: 'Because water supplies and pasture are held in common by all, it is not in an individual's interest to limit his herd size in an attempt to prevent overgrazing and destruction of the environment. Even if he were aware of the consequences of overgrazing, and few are, he would not limit the size of his herds. To do so would jeopardize his personal survival for the benefit of a collectivity that might not concurrently follow his example' (see also Livingstone, 1977:210). It is no coincidence that in all nomadic societies, to a greater or lesser degree, there are inequalities in the ownership of livestock. 18

As for short-term maximization, apart once again from the aims and achievements of separate individuals and their households, essentially this results not in increased production but, on the contrary, in maintenance of the homoeostasis by, it is true, very specific methods. Increased production in nomadic society as a whole is, to any significant extent, impossible; what this really means in the long term is that pastoral economy is doomed to stagnation. Dahl and Hjort (1976:271) are right to point out that: 'The opportunities of building up large herds solely by means of biological reproduction are generally more limited than is implied in many references' (cf. Nicolaisen, 1963:117 on the Tuareg). Thus when Paine (1971:170) writes that pastoralists 'look like rudimentary capitalists' he is following the 18

The question of whether or not nomads themselves understand the impossibility of long-term and unlimited maximization of numbers of livestock needs more research. T h e materials which I have to hand suggest they do not acknowledge, or do not completely acknowledge it, although exceptions to this rule have been recorded.

76

Problems

of balance and non-autarky

of the pastoral

economy

traditional notions of the Formalist school, rather than reality. The outwardly superficial similarity which sometimes exists cannot disguise the qualitative difference between the aims and results of production in a pastoral nomadic society, on the one hand, and in a capitalist one, on the other. However, even an outward similarity between these two completely different types of economics cannot always be found. Amongst the Barabaig the introduction, on the initiative of colonial powers, of a veterinary service led to an increase in the number of livestock and, as a result, to overgrazing and the exhaustion of pastures. Hopes that surplus stock would come on the market did not materialize; the Barabaig's utilization of this stock was traditional - to acquire new wives and to strengthen social standing. The usual explanation for this is the 'cattle-complex' characteristics of East African pastoralists (Klima, 1970:109-10), a striving towards unlimited herd expansion, which cannot be sustained by the economic resource base of the society. In this respect a statement of Sir Philip Mitchell quoted by Allan (1965:331) is indicative: 'There are only two alternatives for the African people - either they eat their surplus stock or their surplus stock will eat them.' With certain variations this statement remains apposite today (see, for example, Whyte, 1966:345). But is it really the case that a cattle-complex leads to unlimited herd expansion and is not a response to economic necessity? If the economy of East African nomads is not regarded as a model of a contemporary capitalist market economy, then it is clear that although a cattle-complex cannot be explained by economic expediency alone (social factors and the value system of East African nomads also must be taken into account), it cannot be considered to be economically irrational. Periodic droughts, epizooty and other natural disasters are the constant companion of the nomadic economy and they entail large-scale losses of stock. Consequently, unlimited herd expansion is impossible (see, for example, Gulliver, 1955:164 on the Turkana or Spencer, 1965:25 on the Samburu). Other factors must also be taken into account. These are the slow rate at which cattle reproduce (Spooner, 1973:11-12), and the desire to prolong the lactation period when milk is obtained, which depends on an increase in the number of young animals (Stenning, 1965:373). In addition, it is in periods of drought when there is insufficient milk that animals are most often sacrificed, which means a source of meat is opened up (Schneider, 1957:278-300). The sacrifice of livestock in this way is linked to institutions of reciprocation necessary to the society. 19

" Ai) interesting comparison is to be found in research on Shanklin (Cramer, 1976), in which it is shown that the seemingly irrational notions connected with cattle amongst Irish pastoralists really have an adaptive function. They make possible the preservation of natural resources in a traditional economic system.

77

Nomadism

as a distinct form of food-producing

economy

Finally, the price of livestock which sometimes is not favourable for East African pastoralists (on the Karimojong see R. and N. Dyson-Hudson, 1969:82) makes it difficult for them to participate in the market economy. In these circumstances the striving towards maximum herd expansion in favourable years is perhaps economically justifiable, as it is an insurance policy for when unavoidable natural disasters occur (Deshler, 1965:167; R. and N. Dyson-Hudson, 1969; Spencer, 1965:4; Spencer, 1973:180; Baker, 1975:191). This point of view has been thoroughly substantiated in some very detailed research carried out by Dahl and Hjort (1976:129; see also Dahl and Hjort, 1979:12ff.). In particular they have shown that the consequences of natural disasters can adversely affect the age- and sex-composition of herds of large stock as much as twenty years after the disaster occurred. But natural disasters and correspondingly sharp fluctuations in the number of stock in herds are characteristic of all nomads, not only of nomads in East Africa. The economic problems faced by all nomads are, essentially, one and the same. Wherever the solution to these problems basically has to rely on the resources of its own economy, the choice of corresponding methods is also limited. In this way it is not only long-term increased production which is incompatible with pastoral nomadism; a homoeostatic balance in the economy of the latter is only attained with great difficulty. It is no coincidence that amongst many nomads a striving to increase their production base by direct utilization of the products of nature may be observed, meaning that hunting, gathering and even fishing are widespread amongst these nomads as supplementary forms of economic activity (see, for example, on the reindeer-herders of the North: Fjellstrom, 1964 and Krupnik, 1976; on the Mongols: Formozov, 1928:142-3 and Potanin, 1950:130; on the Kirghiz: Aitbaev, 1959:77-112 and Shibaeva, 1973:102; on the Kazakhs: Viatkin, 1947:70-1; on the Kalmucks: Zhitetsky, 1892:37; on the Turkmen: Bregel, 1961:49; on the Behmai: Fazel, 1973:137; on the Lurs: Hole, 1978:140; on the Baluch, Salzman, 1972:187 and Spooner, 1975:178; on the nomads of Afghanistan: Ferdinand, 1969:147; on the Rwala and other nomads of Arabia; Musil, 1928:15, 95 and Pershits, 1961:31; on the Tuareg: Capot-Rey, 1962:303 and Nicolaisen, 1963:170, 179-80 and Smith, 1978:77-83; on the Teda: Cline, 1950:29 and Chappelle, 1958:193; on the Somali: Lewis, 1955:73; on the Karimojong: R. and N. Dyson-Hudson, 1969:78, 79; on the Dodos: Deshler, 1965:166-7; on the Barabaig: Klima, 1970:14). Consequently, the division suggested by Salzman (1971) of pastoral nomadism into the pure and multi-resource varieties is fairly conditional. To a certain extent all variants of pastoral nomadism are of a multi-resource kind (cf. Forde, 1963:405). 78

Problems

of balance and non-autarky

of the pastoral

economy

However, food-extracting economies in the capacity of supplementary economic activity cannot, of course, solve the essence of the problem - the instability and one-sidedness of pastoral nomadism. For example, amongst the Tuareg one household can gather several hundred kilograms of wild seeds in a year; but this is insufficient, especially in view of the fact that in dry years when the demand for vegetable food is particularly great, the harvest of wild plants is significantly smaller (Nicolaisen, 1963:179). A possible temporary way out of this situation has consisted in increasing the production base by opening up new pastures or turning agricultural territories into pastures. To a certain extent nomads are able to increase their ecological zone. However, in order to do this a number of propitious factors, which by no means are all of an environmental nature, are necessary. Enough examples of such situations are to be found in the history of ancient times and the Middle Ages. In the first centuries A . D . the Sarmatian tribes of Iazygians and Roxolans, and later on the Huns, occupied the fertile plains of Pannonia (Harmatta, 1970:26ff.). Under the Samanids, some groups of the Oghuz who had left their homeland got permission from the government to occupy some land in Maveraunnahr which was under pasture as it was unsuitable as agricultural land (Barthold, 1963:317). The Mongol conquests led to a very marked decrease in the amount of land cultivated in the lands they subdued and, correspondingly, to an increase in the amount of pastureland. After the subjugation of North China, many Mongols together with their families and livestock settled in North China and Manchuria (Vladimirtsov, 1934:125). In the words of Rashid al-Din (1952a:279): 'To some other of those troops he [Jenghiz Khan] gave places without measure on the borders of Khitai, Jurche and in the borderlands of Mongolia which touch those boundaries. These troops were very numerous, they took possession of all the steppes, towns, winter pastures and summer pastures in Khitai, Jurche and Mongolia and settled down there.' In the pre-Mongol period Semirechye was a well-organized agricultural area in which urban life was quite developed. It suffered relatively little from Mongol invasions as its submission was voluntary. Nevertheless, a few decades later the towns of Semirechye were for the most part in ruins and the agriculture of the area greatly diminished. The explanation for this is very simple. The rich pastures in the area attracted Mongol nomads with all the ensuing consequences (Barthold, 1963b:153). In Iraq the Mongols destroyed the irrigation-network connecting the Tigris to the Euphrates, turning ancient cultural lands back into desert (Coon, 1976:148). (Incidentally, it has been maintained that the destructive 20

J

" Khitai ••• North China " Jilrchc - territory of the JUrchcn in contemporary Manchuria.

79

21

Nomadism

as a distinct form of food-producing

economy

consequences of the Mongol invasion on the irrigation network in Iraq have been exaggerated - see Fernea, 1970:25-37). The fertile lands of Jazira in the north of the Fertile Crescent were deserted by agriculturalists from the time of the Mongol invasion and remained in the hands of nomads right up to the twenties of the twentieth century (Smilianskaia, 1979:25). The Mongol invasion, and the marches of Timur with the intestine wars which followed on from them destroyed the artificial irrigation system on the territory of what is today Afghanistan. The lands of the ancient agricultural oases of Qandahar, Kabul and Peshawar were turned into waste lands, suitable only for pasturing livestock (Reisner, 1954:31). In North Africa, from the time of the Vandals, nomads have gradually pushed out agriculturalists. Particularly destructive was the invasion in the eleventh century of the nomads of the Banu Hilal and the Banu Sulaym, who pushed agricultural life into the land by the sea and into the mountains (see, for example, Julien, 1956a: 145; Levtzion, 1977:360; cf. Bovill, 1958:4, 5-8; see also pp. 108, 287-8 of this book). In the last three hundred years the Chukchi have increased their pasturing territory considerably, to the cost of their neighbours (Bogoras, 1904:15, 732; Leeds, 1965:87-90). At the end of the seventeenth century, the Nentsy began to open up the Yamal peninsula and to push the Entsy out of the basin of the middle and lower reaches of the River Taz and the left bank of the Yenisei; and the Komi crossed the Urals in the nineteenth century and seized pastures to the west of the Ob (Vasilev, 1976:326, 334). Over the course of many centuries the Fulani have spread, together with their livestock, all over the Sudano-Sahelian zone (Stenning, 1957). However, the method described above for increasing grazing territory also could never be constant and stable, if only for the reason that it was too dependent on the balance of forces. Examples are well known of how agriculturalists, when they became stronger than nomads, drove the nomads from the lands they had seized so that once again pastureland was turned into cultivated fields. In the present day this process predominates almost everywhere (see, for example, Le Houerou, 1976:269; Gallais, 1977:306ff.; Dahl and Hjort, 1979:5ff. on Africa). Moreover, by no means all the lands which nomads have been able to seize by force have made suitable pasture. Nomadic reindeer-herders could not open up the taiga, or African cattle-breeders the tropical forests, or Eurasian steppe-dwellers the forest and partially even the forest-steppe zone. The territory of the distribution of pastoral nomadism is more than anything else determined by landscape zonality. As Sahlins and Service (1960) have noted, sometimes less specialized cultures are more capable of making changes than highly specialized ones. However, most importantly, by increasing their ecological zone nomads find no complete and long-term solution to the problem of balance, for 80

Problems

of balance and non-autarky

of the pastoral

economy

sooner or later, even when enlarged, the zone becomes congested. A dynamic homoeostasis, which is achieved with difficulty in a pastoral nomadic economy, in general could not be fully regulated at the expense of inner resources alone, or by their straightforward increase. Despite the still current view to the contrary (see, for example, Bacon, 1954:57; Nicolaisen, 1963:481), pastoral nomadic economies, which I still regard conditionally as autonomous economic systems, on account of their specialization, are not autarkic. When Lattimore (1967:69, 329) writes that a steppe economy can manage on its own better than the economy of sedentary societies can, he does no more than repeat notions which are now out of date. Incidentally, it would appear that he has now reappraised his views (cf. Lattimore, 1979:481-3). It is curious that seizure of livestock, both from neighbouring nomadic societies and from agriculturalists, was sometimes an important motive behind the raids and warring of nomads, along with seizure of agricultural products and handicraft-goods. Right from the start of their conquests and throughout them, the Mongols were as eager to seize livestock as they were to seize everything else. After their march against the Tangut in 1205, they drove off huge herds of stock (Rashid al-Din, 1952:144; Rashid al-Din, 1952a: 150). One interesting characteristic of the reindeer-herding economy of the Nentsy and the Chukchi at the beginning of the twentieth century was the low percentage of products from reindeer-herding itself in the food balance, and the dependence of nomads on bought and exchanged goods. The number of deer available could not fully satisfy all the requirements of the nomadic economy. Thus the total output of the Nuvet Chukchi could only with difficulty satisfy the requirements of the nomads themselves, but reindeer-herding did not provide any surplus output to be used in exchange operations. Consequently, for the Chukchi who herded reindeer, the exchange of reindeer-meat for the meat and, especially, the fat of maritime game was necessary to substitute their own output with the more calorific products of non-nomadic societies, just as the European Nentsy had to purchase foodstuffs. It was the reindeer-herders rather than maritime hunters who were more dependent on exchange operations, and they had to give products of their reindeer-herding (which they themselves needed) for (he fat of maritime game, the surplus of the hunter's catch (Krupnik, 1976:87-8). The Somali often had to sell livestock in the dry season in order to remain alive (Lewis, 1961:43). The Barabaig exchange livestock every year during the dry months, when there is not enough milk to go round, for maize and millet from neighbouring agriculturalists (Klima, 1970:13, 25). In years of drought the Tuareg feed almost exclusively on agricultural products (Nicolaisen, 1963:200). HI

Nomadism

as a distinct form of food-producing

economy

A nomadic economy also limits opportunities for division of labour, especially the development of handicraft as an independent branch of the economy. My aim here is not to analyse the economic interrelations between nomads, agriculturalists and, especially, townsmen in all their complexity. This subject will be discussed at great length below. At present I should like to emphasize just one side of these reciprocal relations, the side which Ibn Khaldun (1967:122) was already able to describe with great lucidity:'. . . the desert civilization is inferior to urban civilization, because not all the necessities of civilization are to be found among the people of the desert. . . While they [the Bedouins] need cities for their necessities of life, the urban population needs [the Bedouins] for convenience and luxuries. Thus, as long as they live in the desert and have not acquired royal authority and control of the cities, the Bedouins need the inhabitants of the latter.' The fact is that specialists have noted the non-autarky of nomads in regions they are studying, sometimes mistakenly supposing that things are different in other regions. Thus Barth (1962:345) records the dependence of nomads in Southwest Asia on the agricultural products and handicraft goods of sedentary societies, and writes that reindeer-herders in the North and nomads in Central Asia are almost fully independent economically of the outside world. His first assertion is completely true, the second, alas, is not. In effect, examples of the non-autarky of nomads in the Middle East are only too numerous and well known (see, for example, Schurmann, 1962:261; Ferdinand, 1969:136ff.; Swidler, 1973:23; Spooner, 1975:173). And the same has been true of other nomads, at least of the overwhelming majority of them. Possibly quantitative differences have existed, but not qualitative ones. I have already written about the nomads of North Eurasia. The nomads of the Eurasian steppes, despite the opinion of certain scholars, have also been quite unable to manage on their own. Mahmud Kashgari had long ago introduced the saying: 'There is no Turk [i.e. nomad - A.K.] without a Tadjik [sedentary], there is no hat without a head' (quoted in Barthold, 1963d:460). The autumn raids of the Polovtsians (the Cumans) who hoped to seize the fruits of the harvest were particularly dangerous for the Russians (Kargalov, 1967:37). The raids of nomadic Uzbeks and Kazakhs on the agricultural region of Maveraunnahr occurred almost exclusively in the winter months, for it was in these months that they had insufficient fodder for their livestock and experienced the great­ est hardships. According to Fadlullah ibn Ruzbikhan, Kazakhs launched raids on towns subordinate to Shaybani Khan, '. . . as a consequence of their great need of clothing and cotton garments' (Akhmedov, 1965:82, 110). The nomads of the Near East have been just as strongly dependent on agriculturalists and craftsmen. I shall limit myself here to concurring with the opinion of authoritative specialists (Hitti, 1956:19; Capot-Rey, 82

Problems

of balance and non-autarky

of the pastoral

economy

1953:212; Capot-Rey, 1962:302; Nicolaisen, 1963:209-10, 213-14, 216, 481-3; Rosenfeld, 1965:75; de Planhol, 1966:278). The notion that East African and Sahelian pastoralists are autarkic and independent of agriculturalists is quite often asserted, but it too is in need of reassessment (see, for example, Gulliver, 1955:6-7; Stenning, 1965:365-72; Deshler, 1965:166; Winans, 1965:173-5; July, 1975:157). In this way, economic instability and non-autarky can be considered as almost indispensable attributes of a pastoral nomadic economy. Nomads have had two alternative ways of overcoming them. (1) Sedentarization which, if it affects an entire nomadic society, is usually linked to a break-up of the basic structure of the society. However, in an ecological zone occupied by nomads sedentarization has almost never been fully practicable (at least in the past it has never been practicable). Consequently, in practice sedentarization usually has been realized at best as one potential tendency of development. In so far as nomadism developed as an alternative to agriculture in just those regions where agriculture was economically unprofitable, the seden­ tarization of nomads demanded either that they migrate into other regions (or at least they moved into other ecological niches or enclaves - on this point see pp. 199f.), or it was an enforced measure. Only where agriculture was a possible alternative, particularly where it had economic advantages over pastoral nomadism, did real sedentarization take place. Looked at from this angle, semi-nomadism is not always a step on the path towards complete sedentarization or the optimum economic variant in its ecological zone. Often it is not so much a spatial as a temporal alternative to full nomadism, when the latter cannot guarantee to provide the necessary products from outside in the required quantity and/or on terms which are profitable to itself and, therefore, has to search for internal resources. Such enforced sedentarization which is linked to the instability of nomadic economies in reality affects primarily the poorest nomads. It would appear that sedentarization as one possible consequence of the impoverishment of nomads is a phenomenon characteristic of all nomads, and one which has frequently been condemned and generally despised by nomads. I have written about semi-nomadism amongst the northern reindeerherders of Eurasia (see pp. 41-2). Already in the eighteenth century impoverished Lapps were settling in valleys where they engaged in fishing and hunting (Vorren, 1973:146). However, since more often than not it was linked with the impoverishment of individual households, in the North semi-nomadism was a temporary state rather than a stable system. In the Eurasian steppes, such sedentarization has been known since Scythian times (Khaznnov, 1975:148-9). The Scythians looked upon the impoverished who could no longer roam from place to place as 83

Nomadism

as a distinct form of food-producing

economy

'dishonourable people . . . of the lowest provenance.' Amongst the Oghuz, those who had become impoverished and settled on the land ceased to enjoy full rights as members of society. The common law of the Oghuz looked benevolently only on the nomadic way of life. The following statement is attributed to the legendary Oghuz Khan: 'Move about, do not be sedentary, roam in spring, summer and winter pastures and lands by the sea, wanting for nothing. Let not your supplies of milk, yoghourt and kymran be decreased' (quoted in Agadzhanov, 1969:96-7, 109). Barth (1962:350; 1964a:77-8) also records that amongst the nomads of Southwest Asia impoverishment is the most common path to sedentarization. The same is true of Africa today (Gallais, 1977:310ff.; Dahl and Hjort, 1979:28ff.). Enforced sedentarization has by no means always been so stable and protracted. Individuals forced to abandon nomadism have regarded their situation as temporary, and have reverted to nomadism at the first possible opportunity. One Russian observer of the nineteenth century wrote of the Kazakhs: 'Only hopeless poverty can rouse the nomad to till the soil. But as soon as he has provided himself with stock, he immediately throws away the clumsy spade he used to till the soil instead of a plough - he becomes nomadic' (Zavadsky-Krasnopolsky, 1874:17). Such sedentarization, as it has been correctly pointed out by Barth (1976:76) and by Dahl and Hjort (1979:28-30), eased the pressure of the population and, indirectly, of the livestock on natural resources. However, only rarely of course did it fully solve the economic problems of the nomadic society. Consequently, nomads often tried to develop agricultural and handicraft sectors in their economy with the help of immigrants from other societies. However, the latter case will be examined below as it is connected not only with economic, but also with political problems. (2) The second alternative consists of acquisition by different methods of needed products from neighbouring sedentary societies. The latter have acted as an additional factor to guarantee the stability of pastoral economies. Thus the adaptation of nomadic economy to its environment is incomplete. The alternative is the necessity and inevitability of the adaptation of nomadism to the outside world. In this chapter I have tried to argue the economic reasons why such adaptation is necessary amongst nomads. However, it is also necessary for social and political reasons. 22

2 2

Boiled cow's-milk, mixed with water or koumiss.

84

2 The origins of pastoral nomadism

Spooner (1973:5) has written that '. . . virtually nothing is yet known about the beginnings of nomadism.' I am not so pessimistic. Of course, it is now impossible to restore the full and detailed picture of the origins of pastoral nomadism, for there are too many missing parts and too many obscurities. However, the general outlines still show through clearly enough. First of all I must say something about contemporary opinions on this subject. Since historiography is not the main aim of this book as a whole, or of the present chapter, it is neither possible nor necessary for me to examine all the numerous theories about the origins of pastoral nomadism, some of which are exclusive, and some of which supplement each other. I shall only pause on some of the more well-known theories. The so-called 'tripartite theory', according to which pastoralism de­ veloped out of hunting and, moreover, earlier than agriculture, for a long time exercised a strong influence on the way in which the problem of the origins of pastoral nomadism was approached. The theory first appeared back in antiquity (Dicaearchus, Varro), but it was in the eighteenth and nineteenth centuries that it became particularly well known. Montesquieu, Herder, Condorcet, Mortillet, Lubbock, Morgan and Engels were all adherents. Admittedly, Vico thought that agriculture emerged earlier than pastoralism, but his opinion did not receive recognition. Admittedly, Tylor, Lippert, Schurtz, Schmidt, Koppers, Menghin and Lowie assumed that pastoralism and agriculture could have appeared independently of each other, but they too concluded that pastoralism developed directly out of hunting. Even Hahn (1896; cf. Kramer, 1967), who proved that the domestication of animals occurred only amongst horticulturalists, took some time to convince anthropologists and archaeologists (although by no means all of them) of his opinion. In this connection I must point out that the Bible gives chronological priority not to pastoralism, and certainly not to pastoral nomadism, but to agriculture (cf. Genesis 4.2 and, particularly, Genesis 4.20). Neither Cain, K5

The origins of pastoral

nomadism

nor even his younger brother Abel, 'a keeper of sheep', but only 'Jabal: he was the father of such as dwell in tents and of such as have cattle.' Some adherents of the tripartite theory have worked on the problem of the origins of pastoralism in general, but have specifically been engaged in studying the origins of pastoral nomadism. Such scholars believe that the emergence of the latter was directly linked to the hunting of herbivorous mammals. Wandering hunters who followed herds of these animals finally managed to domesticate them and thus became pastoral nomads. Reindeer-herding usually has been cited and, indeed, still is cited as evidence of this process. Even many of those who only have accepted the tripartite theory with specific reservations, or those who have not accepted it all have regarded reindeer-herding as the most ancient form of pastoralism, going back almost to the paleolithic period (see, for example, Bogoraz-Tan, 1928:83-5; Schmidt and Koppers, 1924:502-28; Schmidt, 1951:1-41; Flor, 1930:86-146,152-237; Thurnwald, 1932:77; Pohlhausen, 1954; Pohlhausen, 1972:176-95; Hermanns, 1949:216f.; Curwen and Hatt, 1953:37-46, 93-4, 278). In all fairness it should be pointed out that in the opinion of some scholars it was not only out of hunting that pastoral nomadism and, particularly, less specialized forms of pastoralism emerged. These scholars have assumed that agriculturalists also could have begun the breeding of animals (see, for example, Lattimore, 1967:327). However, even this compromise point of view invites many objections. Below I shall try to show that the idea, deeply rooted in the minds of many scholars, that reindeer-herding is the oldest form of pastoralism contradicts facts. In the meantime I shall merely point out that although instances of hunters taming individual animals are not infrequently described in ethnographical literature, I regard the idea that animals were domesticated and herded by hunters with scepticism. The mobile way of life of hunters, their lack of necessary fodder reserves and many other factors were serious obstacles. Even less were hunters in a position to domesticate an entire herd. Hunters of caribou and wild reindeer were quite unable to follow herds as they could not maintain the necessary speed of movement this required and, moreover, when they were following herds they were unable to engage in any other economic activity. Once they had learnt the migration routes of the deer, they hunted deer at a specific time of year and in specific places (Burch, 1972:344-59). Moreover, the changeable composition of the herds made their domestication impossible. The most that hunters could do was borrow animals which already had been domesticated; some reindeerherders are an example of this process. This, evidently, was one of the minor ways in which pastoral nomadism spread in the oikoumene, and was not very common. In this connection one fact springs to mind, this is that the 86

The origins of pastoral

nomadism

American Indians of the Great Plains became mounted hunters of bison herds, but in no way pastoralists and nomads. Some scholars attach great significance to climatic changes connected with desiccation (Toynbee, 1935:8, Zeuner, 1956:28). It is now an established fact that in the past climates were changeable; but when, how and where they changed remains an area of hypotheses and discussion. Specialists accept that in practice there exists no generally accepted theory of climatic changes (Sutcliffe, 1963:277; Butzer and Twidale, 1966:135; Budyko, 1974:5), although to date theories which attempt to explain climatic variations of differing duration would appear to supplement, rather than to contradict each other. Goethe said that what we find between opposing points of view is not truth, but the problem. But sometimes both points of view contain part of the truth. It is now acknowledged that climate is influenced by many different factors: the earth's orbital variations (Milankovich's theory), variations of tidal forming-forces (Petersson's theory), solar output variations, tilts in the earth's axis, continental drifts, etc. (Brooks, 1949; Bell, 1953:123-36; Shapley, 1953; Fairbridge, 1961; Lamb, 1966:3ff.; Shnitnikov, 1969; Vozovik, 1970). Acting both separately and concomitantly, these factors create a complex picture of climatic oscillations of long and short duration. Unfortunately, it is particularly difficult for paleoclimatologists to establish the chronology, and sometimes even the actual presence of short fluctuations; however, both these factors are of particular interest for the problem being examined here. For example, stability of zonal borders (particularly between forest and steppe) does not always indicate stability of climate, because forest boundaries are markedly unchanging, even in dry conditions, and spread relatively slowly into new areas, even in favourable humid conditions (Milkov, 1967:137; Lamb, 1974:196-7). However, annuals, both wild and cultivated, are not forest; for them a small temperature change and, particularly, small changes in humidity can have very important consequences. In so far as I know the majority of paleoclimatologists now refute the ideas of Huntington and his followers (Huntington, 1907; Huntington, 1915; Huntington, 1945; Grumm-Grzhimailo, 1933; Petrov, 1966a:205-ll; cf. Chappell, 1971) about pulsatory desiccation of climate, along with assumptions about the practical immutability of climate in the Holocene (Raikes, 1967:92). As Brooks (1949:286) has written,'. . .the progressive theory offers the opportunity for a definite negation, while the theory of fluctuations does not.' Over the course of the last ten to twelve thousand years fluctuations of climate, both in the direction of aridity and of humidity, both local and planetary, both more or less prolonged and of very short duration, have frequently occurred. In no way would I wish to deny that some of these could have affected the destiny of societies with food87

The origins of pastoral

nomadism

producing economies, pastoral societies amongst them, in a most essential way- On the contrary. However, dry periods occurred repeatedly, but the emergence of different types of pastoral nomadism in each of the regions of its distribution was not a simultaneous or short-term event, although it decisively occurred only once. Consequently, in contrast to Lattimore (1967:158), I consider that climatic changes played an important role in the emergence, if not of all, at least of many types of pastoral nomadism; nevertheless, I consider that this alone is insufficient explanation. Economic and cultural preconditions were also necessary. Fairly popular, particularly in Soviet literature on the subject, is the idea according to which the transition to pastoral nomadism was linked to increase in number of head of livestock in herds and to difficulties in feeding them. For this reason pastures had to be changed periodically, and livestock was pastured further and further away from settlements. In the end those groups which were interested primarily in pastoralism ceased to engage in agriculture, even as a secondary occupation, and became nomads (Sauer, 1952:97; Hancar, 1955:556; Griaznov, 1955:19-29; Griaznov, 1957:21-8; Bogoliubsky, 1959:174, 175; Gumilev, 1966:67-8; Artamonov, 1977:6; for criticism of this view see Markov, 1976:22-3). This suggestion, which is purely evolutionary in the tradition of the nineteenth century, attributes decisive significance to slow, smooth quantitative changes; but it is quite open to criticism. Firstly, as much ethnographical data shows, pastoralists, even the most primitive amongst them, were able to regulate the size of their herds without impairing the latters' reproductive capacity, and resorted to doing this whenever necessary. Secondly, the number of animals in nature also is regulated, although we do not fully understand how. What is clear, however, is that regulating factors function more effectively when density of animal population is high than when it is low, and when the practicable balance with food resources is upset (Lack, 1954). We can assume that regulating mechanisms also affected the most ancient domesticated species of animals, which differed little from their wild ancestors. Thirdly, as I have already noted in the preceding chapter, in conditions of extensive pastoralism (and not only nomadic ones) prolonged and ceaseless increase in numbers of livestock in herds is impossible. Initial increases in numbers in herds were linked to the opening up of new ecological zones, and they were a consequence rather than a cause of transition to pastoral nomadism. The other side of this advantage was a whole series of shortcomings, which resulted from one-sided economic specialization. Consequently, it is 1

1

Admittedly, Zeuner (1956:28) surmises that the horse and camel were domesticated by tribes which earlier had made the transition to nomadism because of climatic desiccation. However, contemporary data refutes this suggestion.

88

The origins of pastoral

nomadism

doubtful that aspirations to increase numbers of stock were a major motivating force of nomadization. It would appear that yet another idea about the emergence of nomadism can be traced directly or indirectly back to Lattimore (1967:277-8). Essentially his views relate to a particular case - the emergence of nomadism in Inner Asia, which Lattimore has explained as the result of pressure exerted by agricultural societies to push out their weaker neighbours into the steppes and deserts. Below I shall try to show that even with regard to Inner Asia Lattimore's point is debatable. But amongst certain scholars (see, for example, Spooner, 1975:5, 6, 40; Lees and Bates, 1974:187-93; Service, 1975:29-50; Gilbert, 1975:70; cf. Khlopin, 1970:55; Shilov, 1964:102) it has taken on a more generalized form and become a 'theory of displacement' in its own right. The emergence of pastoral nomadism is explained as the result of several groups being pushed into arid zones, either because of the pressure of stronger neighbours, or because of overpopulation, or because of both these reasons at the same time. In my view this rather speculative hypothesis is not substantiated by facts, at least in general it is not. The majority of steppe and desert areas into which separate groups were supposedly pushed had already been settled, and, moreover, by groups of people involved in food-producing economies. It would appear that the emergence of pastoral nomadism was so complex and multi-faceted a phenomenon that it cannot possibly be explained by any one isolated factor. Furthermore, in this process local particularities must be separated from universal tendencies and regularities. No less complex and multi-faceted a phenomenon was the spread of nomadism in the oikoumene, namely its adaptation and re-adaptation in different ecological zones, specific social transformation, the interaction and inter-influence of different forms of pastoralism and other kinds and forms of economic activity. The sources of pastoral nomadism are now more or less clear. They go back to the Neolithic revolution and to the emergence of food-producing economy which, it is now clear, in the Old World had always basically consisted of two forms of economic activity - cultivation and animal husbandry. It is not impossible that groups involved in advanced gathering had already begun to domesticate animals before incipient cultivators did. What is most important, however, is that only groups leading a relatively sedentary way of life and who had definite surpluses of vegetable food at their disposal could domesticate animals. Food-producing economies evi­ dently spread from the centres of their initial emergence in different ways, the most important of which were migrations and borrowings. However, in all circumstances, a food-producing economy entailed adaptation to a different habitat, sometimes the domestication of new species of animals and specific economic specialization, which led in a number of areas (chiefly 89

The origins of pastoral

nomadism

in arid zones) to the predominance of pastoralism over other forms of agriculture. It should not be assumed that this process was an automatic one. Indeed, the more mobile forms of pastoralism which emerged from more sedentary forms did not always completely crowd out the latter. Often the two co-existed, and this was one factor in the further diversification and specialization of pastoral economy. Thus, in some areas initial household animal husbandry with free-range grazing led to the appearance of more developed variants of sedentary pastoralism, and in other areas to herdsman husbandry. Where there were suitable conditions and stimuli herdsman husbandry in its own turn could be an initial point of departure for the emergence, first of semi-nomadic pastoralism, and then of pastoral nomadism. Of course, we can only talk of such developments in areas in which the development of the cited forms was basically spontaneous. Moreover, nomadization itself was never anywhere absolutely predeter­ mined; everywhere it developed it had its own preconditions, stimuli and motivating and inhibiting factors. The origins and spread of pastoral nomadism in its principal and secondary centres will now be examined more specifically. The Eurasian steppes, semi-deserts and deserts. The emergence of food-producing economy and changes within it in the direction of pastoral economy did not occur simultaneously. This took several thousand years. The process was much more complex than the way in which it is presented by Goodenough (1970), who links it with the advance of 'Kurgan culture' from the West into the steppe. However, this culture, which Gimbutas (1970:155-97) insists really existed, is only an artificial and speculative construction which unites under one heading many archaeological cultures which themselves are very different and are from different periods. Equally untenable is the opinion (Bacon, 1954:49-51; see also Vainshtein, 1978:130) that hunters who borrowed livestock from neighbouring agriculturalists were the first nomads in the Eurasian steppes. Food-producing economy penetrated the East European steppes from the Balkans as far back as the sixth and fifth millennia B.C. (Passek and Chernysh, 1970; Formozov, 1977:63-79). In the sixth millennium B.C. agricultural cultures of South Turkmenia were directly linked with early agricultural cultures in Western Asia and for a long time formed a peninsula in a sea of cultures in which there were food-extracting economies. Traces of pastoralism are to be found only in the upper levels of Kelteminar culture of the third millennium B.C. (Vinogradov, 1968; Formozov, 1977:119). The northern steppe zone of Middle Asia knew nothing of food-producing economy until the second half of the third millennium B.C. 2

2

Kurgan - a Russian term borrowed from the Turkic languages for burial mound, tumulus,

90

The origins of pastoral

nomadism

Further east, in South Siberia, the first appearance of a food-producing economy can be traced only in Afanasevo culture in the second half of the third millennium B.C. (Istoriia Sibiri, 1968:159-62). The Neolithic inhabi­ tants of Mongolia and the Trans-Baikal area, despite the opinion of Egami (1970:324), were not nomads; they were hunters and gatherers (Volkov, 1967:90; Okladnikov, 1970:179). Only in the second half of the third and the second millennia B.C., in the Bronze Age, did food-producing economy (in several different variants) finally become predominant all over the Eurasian steppes. Already in the fifth and fourth millennia B.C. in the southern areas of East Europe domestic cattle and small stock, and even horses were known (Tsalkin, 1970:265). Already on the border of the fourth and third millennia B.C. in certain areas of the European steppes, particularly in the area between the Volga and the Urals, the predominance of pastoralism over agriculture was to be observed (Zbenovich, 1974:112-14; Merpert, 1974:102-12). The spread of pastoralism in the Neolithic and Bronze Ages was also characteristic of different areas of Central and Western Europe (see, for example, Fleming, 1972; Bradley, 1972). However, despite the opinion of certain scholars (Merpert, 1974:112; Shilov, 1964:101, 102; Shilov, 1975:14; Masson, 1976:44), there are no grounds for thinking that pastoralists of the third and even the second millennia B.C. were real nomads. Judging from osteological materials, in the Neolithic, Eneolithic (Chalcolithic), and Bronze Ages there was no significant change in the Eurasian steppes in the species-composition of herds and the percentage correlation of the different species within them, in the directions which later on became characteristic of nomads in this region (Tsalkin, 1970:253). In many archaeological cultures in this period long-term settlements were the norm, a fact which also testifies against the presence of pastoral nomadism in the Eurasian steppes at this time. Finally, it is difficult to imagine nomads of the Eurasian steppes not using the horse as a riding animal. Thus, although the problem of the first use to which the domesticated horse was put and its ramifications is fairly confused, in this context it needs to be examined. In South Russia the horse was domesticated very early, no later than the fourth millennium B.C. It could be thought that at that time the horse was not yet of any great importance to the economy; however, there are two instances which demand explanation. On two fourth millennium B . C . sites (or third millennium B . C , depending on the chronological system being accepted), Dereivka, in the forest-steppe zone and Repin Khutor, in the Lower Don steppe, bones of horses constitute the majority, up to 80%, of osteological remains of domestic animals. In this connection Tsalkin (1970:248) has written about 'the first tribes of horse-breeders in the history of animal husbandry'.

The origins of pastoral

nomadism

But what was the practice for pasturing these horses? We have no incontestable and positive data about riding in the fourth millennium B.C. At the same time it is hard to imagine herds of horses which, moreover, had only recently been domesticated, being managed by people on foot. I can see only one explanation for this problem. We must return once again to the old problem of Fahren und Reiten'. According to the point of view, which to date is most widespread, the horse was first captured to be used as a draught animal, only after this did it become a riding animal. However, other scholars insist that it was the other way round (see, for example, Curwen and Hatt, 1953:47; Trubachev, 1960:15; Kovalevskaia, 1977:21). Because there is no data confirming that animaldriven wheeled transport appeared before the third millennium B .c., it is not impossible that those scholars in part are right. But they are right only in part. This is because riding, if it did exist in the fourth and third millennia B.C. and even later on, had to remain undeveloped. Either there were no harnesses at all for the horses, or only the most primitive ones. In some art of the second millennium B.C. from Western Asia, people are depicted sitting on horses without saddles and bridles. But they were not yet real horsemen; at most they were the forefathers of such. This was not only because they were little able to engage in warring activities (Kussmaul, 1963:31ff.) - even their pastoral possibili­ ties were limited. Moreover, they were an isolated, rather than a mass phenomenon. The pasturing of small herds of horses with the help of dogs, sometimes on horseback and without going too far away from the settlement is one thing. The numerous pig-bones on the Dereivka site indicate a predominantly sedentary way of life. But quite another thing is the regular following of herds on horseback along routes of hundreds of kilometres. According to the data which at present may be considered trustworthy, the appearance of the first real horsemen occurred around the middle of the second millennium B . C . (Smirnov, 1961:46; Zeuner, 1963:337). In the second millennium B.C. there existed in the Eurasian steppes so-called 'steppe-bronze cultures'. Although the correlation of pastoralism and agriculture in these cultures and even between different local variants within individual cultures varied, the economies of all of them were complex pastoral-agricultural or sometimes even agricultural-pastoral ones. At any rate in no way were they nomadic (Akishev and Kushaev, 1963:131; Tsalkin, 1964:29; Tsalkin, 1972:66-81; Istoriia Sibiri, 1968:172-3; Itina, 1977:39). At the same time a series of data testifies to the fact that in many steppe cultures pastoralism at an early date assumed extensive forms. It would appear that in the Bronze Age and even in the Eneolithic (Chalcolithic) Age the inhabitants of the steppes were already opening up river valleys and l

92

The origins of pastoral

nomadism

penetrating quite deep into the steppe. Between the Urals and Volga burial grounds between 15 and 90 kilometres away from river valleys have been excavated, and in these grounds not only men were buried, but also women and children (Shilov, 1964; Merpert, 1974:98-100). It would appear that in the third millennium B.C. the steppe was more humid than it was later on. But there is another possibility. This is that some families and even groups were already breaking away from the stationary settlements for sizeable amounts of time. The opening up of the steppe was made easier for pastoralists in the Bronze Age by the fact that the knowledge of animaldriven wheeled transport came to East Europe from Western Asia no later than the third millennium B . C . Thus it could be thought that in the third and second millennia B.C. in the steppe there were pastoralists whose way of life was mobile and who herded sheep and, to a lesser degree, large stock which they followed on foot or in carts drawn by oxen or horses, or sometimes even on horseback. In the different regions of the steppe their pastoralism assumed different forms, the most mobile of which were evidently herdsman husbandry and in places semi-nomadic pastoralism even. It is even possible that there were some groups which had nothing at all to do with agriculture, but their existence is still something of a conjecture. But if they did indeed exist then they were different from nomads of later periods, because they were a component part of societies with complex pastoral-agricultural economies. Kroeber (1948:278) has written of nomads that they form not a culture, but a half-culture or part-culture, which depends on sedentary neighbours. With regard to the pastoralists of the Eurasian steppes in the Bronze Age even this is an exaggeration, because their cultures were identical to the corresponding cultures of sedentary agriculturalists. The relative mobility of pastoralists of the Bronze Age in the Eurasian steppes facilitated their migrations which, evidence indicates, were already taking place in the fourth, third and second millennia B.C. Many specialists now place the ancestral homeland of the Indo-Iranians in the southern Russian steppes. Then the appearance of the Indo-Iranians in Western Asia, Iran and India in the second millennium B.C. must also be included in the list of pastoral migrations, which becomes even longer if we take into account the numerous migrations of pastoralists within the steppe itself and on the borderlands closest to the steppe; although these migrations are nowhere recorded in written sources, they can easily be discerned by archaeological data (see, for example, Zbenovich, 1974:161; Itina, 1977:188, 234-5; Merpert, 1978:62). Reasons for these migrations must have been fairly diverse: population increase; exhaustion of pastures, which could have happened fairly quickly 3

1

Linguistic materials permit us to reconstruct a common Indo-European conception about the worid beyond the grave as a pasture (V. V. Ivanov, 1976:33).

93

The origins of pastoral

nomadism

where a regular, all-year-round cycle of pastoral migrations had not been established; and last, but not least, the urge to move closer to centres of advanced agricultural cultures or civilizations. However, just as a cart does not resemble a horse, the migrations of the third and second millennia B.C. pastoralists who were still involved with agriculture did not resemble the migrations of mounted nomads in later centuries. The former were markedly slower and more gradual, and new lands suitable for cultivation interested the migrators no less than new pastures did. To regard the ancient Indo-Iranians as nomads similar to the Scythians and Mongols (see, for example, Leshnik, 1972:150-66; van Lohuizen-de Leeuw, 1975:16) is a mistake which is explicable in the nineteenth century, but hardly justifiable in the present day. All the necessary technological preconditions for pastoral nomadism: corresponding species-composition of herds; long-term practice of mobile and extensive forms of pastoralism; dairying; animal-driven wheeled transport; and horsemanship appeared in the European and Kazakh steppes no later than the middle of the second millennium B.C. But no corresponding transformation in 'steppe bronze cultures' is to be observed. A noticeable break only occurs on the border of the second and first millennia B.C. and, particularly, at the beginning of the first millennium B.C. when life in the settlements of the preceding period ceases, when new archaeological cultures, which have left explicit traces of riding and nomadism, appear, and when the authors of antiquity call the inhabitants of the steppe 'drinkers of milk' and 'milkers of mares', and a little later by specific names - the Cimmerians, Scythians, Sakas and others. The preconditions of the transition to pastoral nomadism, the whole complex of which was on hand already in the middle of the second millennium B.C, remained unrealized for at least half a millennium. I can suggest only one explanation for this - climatic change and change in the economic and political situation of the region. One of the most complex problems faced by paleoclimatologists of the Holocene is that of the Sub-Boreal Climatic stage. Blyti-Sernander's classification of climate, according to which a xerothermic climate characterized this period, is now criticized by several scholars. The point usually made by these scholars is that there is no data which confirms the simultaneous formation all over Europe of the 'border' horizon (Grenzhorizont), indicating that the formation of peat ceased because of the drying out of marshes. Meanwhile research of peat-bogs, 'the incomparable archives of nature' in the words of one specialist, was the basis on which the scheme in question was devised. Moreover, opponents of Blyti-Sernander's classifica­ tion believe that there is no conclusive evidence that the steppe biota penetrated the forest zone in the Sub-Boreal Climatic stage (Predtechensky, 1957:106-7; Buchinsky, 1957:19; Tattar, 1961:336; Neishtadt and Gudelts, 94

The origins of pastoral

nomadism

1961:14-18; Milkov, 1967:127, 131, 136). However, their opinion is disputed by other scholars who use arguments showing that some drying out of reservoirs and encroachment of the steppe onto forest nevertheless did take place (Kleopov, 1941:231ff.; Berg, 1947:97ff.; Tumadzhanov, 1961:263; Kostin, 1965:69; Riabtseva, 1970:117). Since I am not a specialist in this field I naturally shall not take it upon myself to judge which side is right. For me something else is far more important. Independently of their own theoretical positions and specific argumentation, almost all paleoclimatologists accept that the second millennium B.C. was characterized by a dry climate which, it would appear, was at its driest at the turn of the second and first millennia B.C. (Brooks, 1949:296, 305; Buchinsky, 1957:19; Shnitnikov, 1957; Shnitnikov, 1969; Predtechensky, 1957:98, 106; Tumadzhanov, 1961:262, 263; Veryard, 1963:4; Kostin, 1965:69; Lamb, 1966:6; Starkel, 1966:27; Frengel 1966:113; Butzer and Twidale, 1966:135; Sinitsyn, 1967:150; Riabtseva, 1970:117). Archaeological data also confirms this suggestion (Vinogradov and Mamedov, 1975:251; Piperino and Tosi, 1975:186). The fact that these dates coincide with the period of the emergence of pastoral nomadism, as it has been established by archaeological data and written sources, is scarcely due to chance. It would appear that the dry climate was the final stimulus for pastoralists to abandon agriculture once and for all and become fully nomadic. Moreover, this transition almost coincided with the emergence of sedentary states in the land to the north of the Black Sea, in Middle Asia and in the borderlands of these two areas. The various contacts which they had with these states were of great economic, social and political significance to the ancient nomads. It was no coincidence that the Cimmerians and Scythians announced their appearance on the world arena by marching into Western Asia (Herodotus 1.103-6; IV. 1). It was not so much pressure from agricultural states as their very existence with corresponding opportunities for nomads which facilitated the specialization of the latter. If the transition to nomadism in the eastern part of the Eurasian steppes, in Inner Asia, occurred later than it did in the western part, it was not very much later. Admittedly, according to Lattimore (1967:57-8, 341, 346-9), northern and western barbarians, the Jung and the Ti, who troubled China in the first half of the first millennium B.C. were not only pastoralists, but I hey also engaged in agriculture; he insists that the Chinese of the Chou period did not describe these barbarians as mounted nomads. Lattimore (1967:61; cf. Watson, 1972:140) maintains that horses began to be ridden and real nomads appeared on the borders of China only in the fourth and third centuries B.C. Nevertheless, it would appear that all of this happened considerably earlier. In Yin inscriptions the epithets 'horse' or 'herding-horses' are applied to 95

The origins of pastoral

nomadism

the Ch'iang (Jung) (Kriukov, Sofronov, Cheboksarov, 1978:175). In the Lichi treatise on the Jung and Ti it is written '. . . some of them do not use cereals for food.' It is very probable that the transition to nomadism in Inner Asia was linked to certain pressures and impulses from the West, although Lattimore (1967:162-3) refutes this assumption. In the opinion of Kuo Mo-jo and a number of Soviet sinologists, the Ti tribes which appeared in Chinese history in the seventh century B.C. were 'Scythians', that is nomads who spoke the languages of the Iranian group (Go Mo-zho [Kuo Mo-jo], 1959:434; Kriukov, Sofronov, Cheboksarov, 1978:183-4; Volkov, 1979:132-3). Articles fashioned in the Scythian style, including the famous animal style, have been found in Ordos and even in North China. Archaeological materials and the data of physical anthropology also testify to the fact that in the first millennium B.C. pastoralists from Kazakhstan, Middle Asia and, possibly, from the Altai penetrated Mongolia (Volkov, 1967:95; Jettmar, 1967:134; Mamonova, 1979:207). Lattimore (1967:277-8, 327-8, 343, 349, 364) suggests that the economy of the ancestors of the pastoral nomads of Inner Asia was at first a mixed one when they were living on the borders of China, but that they were pushed out into the steppe and became nomadic as the ancient Chinese put more and more territory under cultivation. In fact the picture was more complex. The nomads not only retreated, they also advanced and transferred to a mixed economy after they had moved onto the territory of North China. Their gradual and further assimilation by the Chinese was possibly more significant than the fact that some were pushed out into the steppe. The origins of pastoral nomadism in the western half of the Eurasian steppes, at the time when pressure from sedentary states was impossible, indirectly testifies against Lattimore's suggestions. Lattimore (1967:385-6) insists that if the barbarian tribes of the North had been real nomads they would have exploited the state of disorganization in China in the fifth and fourth centuries B.C. and invaded Chinese lands, making use of this favourable time. However, firstly, the Chankuo period (fifth to third centuries B.C.) was one in which the economic, military and political strength of China was increased and the unification of China gradually began to take place. In this context it is interesting to note the opinion of Shang Yue (Shan Iue [Shang Yue], 1957:51) that nomadic invasions in fact furthered the tendency towards unification (this is noted by Lattimore in his last work - see 1979:481). Secondly, it was the unification of China which in turn brought to life the strong unification of nomads under the sway of the Hsiung-nu, although of course this occurred somewhat later. Thirdly, the building of walls to safeguard against threats from the North began in China as far back as the fourth and third centuries B.C. Thus even in this period there was danger. 96

The origins

of pastoral

nomadism

But danger was not the only thing. The history of interrelations between China and pastoralists (amongst whom already there were real nomads) must be seen to begin not in the Chankuo period, but at least in the Chou Dynasty (Fan Ven-Lan [Fan Wen-Lan], 1958:55ff.). Then we should recog­ nize that it was the Jung who forced the Chou ruler Ping-Wang to transfer his capital to the East, and the Ti who invaded the Central China Plain and then settled in the territory of the Wei Kingdom and for more than a century continued to attack the Ch'i, Lu, Ch'in and Chang kingdoms and the domains of the Chou emperors. This is not all. For a long time the barbarians not only attacked China with success, but partly under the influence of China they even created their own states - for example, the Jung kingdom of Yichu and the White Ti kingdom of Chungshan (Hsiangii) (Kriukov, Sofronov, Cheboksarov, 1978:174f.). It is only regrettable that too little is known to us about the circumstances of the emergence, the character and particularities of these states. The Near East (Mesopotamia, Arabia, Syria, Palestine). In the first chapter of this book I attempted to justify my singling out of a Near Eastern type of pastoral nomadism. However, in common with all other types, this one did not emerge in finished form, rather it is the product of a prolonged historical development. Its borders have altered more than once over the centuries. Consequently, first of all I shall examine the principal territories in which this type of pastoral nomadism emerged. It is now acknowledged that the origins of pastoralism in the Near East the most ancient cradle of the Neolithic revolution - were directly linked with the establishing there of a food-producing economy (Reed, 1959:162939). It would appear that the understandable, although sometimes superficial enthusiasm for recent, indeed sensational discoveries has led to a situation in which many scholars have begun to overestimate the degree of specialization of pastoralism in Neolithic Western Asia. Thus, Flannery (1965:1254-5; see also Narr, 1959:85; Masson, 1976:39) has suggested that already in the seventh millennium B.C. the pastoralism of the inhabitants of the Zagros Mountains had taken on a yaylag form, and that besides their permanent settlements these people also had seasonal camps in the mountains. Some scholars have even begun to write about pure nomads or semi-nomads in the Neolithic period (Kupper, 1959:113-17; Brentjes, 1968:29). Recent research has demonstrated, however, that yaylag pastoralism in the Zagros Mountains can be dated no earlier than the second h a l f o f the fourth millennium B . C . (Mortensen, 1975:23f., 32-3). However, as yet there is insufficient data for this question to be finally resolved. The idea that pure nomads appeared in the Near East in the Bronze Age, in the third and second (or even the fourth) millennia B . C . , has become fairly widespread (Toynbee, 1935:404; Kupper, 1957; Kupper, 1959:117, 119; 97

The origins of pastoral

nomadism

Gelb, 1961:27ff.), along with the idea that conflicts between nomads and agriculturalists in Mesopotamia and Egypt already began in this period. Lees and Bates (1974:187-93) surmise that the separation of pastoralism and agriculture in regions where there were ancient riverine civilizations depended on development of irrigation, which stimulated economic specialization. Early economic and political links between pastoralists and ancient centres of agricultural civilization are, indeed, a specific feature of the Near East. Nevertheless, there is no reason to look upon the inhabitants of the steppe, the Hyksos, ancient Jews, Amorites, Sutaeans, Arameans and others, as real nomads. Recent research (Kupper, 1957; Henninger, 1969; Haldar, 1971; Klengel, 1972) refutes the old, but still fairly tenacious notion (cf., for example, Renan, 1889:14n.l; Van Ess, 1947:65; Fischer, 1966:123) that the ancient Semites were the primordial nomads. The notion, lent authority by its association with the names of Wellhausen, Robertson Smith and Winckler, that the ancient Jews were like a Bedouin tribe as the latter is described by Burckhardt, Doughty and Lady Blunt, does not correspond to reality (Goiten, 1955:24f.). According to Allbright (1960:205), the patriarchs led the life of semi-nomads, dividing their time between tending livestock and agriculture. It is noteworthy that in the ancient Near East there was not even one term designating nomads (Klengel, 1972:17). The western Semites herded sheep and goats, and cattle to a lesser extent. Their main transport animal was the donkey (Salonen, 1955:52). Perhaps not all sheep- and goat-herders kept donkeys, but it seems to me doubtful that there would have been any groups herding only donkeys (cf. Henninger, 1969:53). The composition of herds suggests that ancient pastoralists were unable to move more than 30 kilometres from sources of water. It is known that they lived in the outlying districts of agricultural areas in the zone where precipitation was between 100 and 250 millimetres a year, and besides pastoralism they were engaged in agriculture, particularly in those areas in which they could get by without irrigation or in which irrigation did not involve any complicated technology (Kupper, 1957:115-16). In common with the migrations of pastoralists in the Eurasian steppes in the third and second millennia B.C., the migrations of pastoralists in the Near East in this period consisted, as a rule, of slow and gradual movement, sometimes infiltration, rather than rapid, sudden conquest, the more so 4

5

4

5

In all fairness it must be pointed out that certain scholars, Kupper, for example, differentiate between pastoralists of this period and later mounted nomads, although they continue to use the same appellation for them, evidently following an old tradition. It is possible, however, that they also lived in the more remote regions of the steppes and deserts. According to some specialists, the climate was less dry in this period and petroglyphs which have been discovered in Central Arabia witness the fact that even there ancient pastoralism was possible.

98

The origins of pastoral

nomadism

because pastoralists in the latter region were the immediate neighbours of agricultural regions. 'But it is less a question of the migration of entire tribes than that of segments of tribes, of smaller groups' (Klengel, 1972:37; cf. Kupper, 1959). Migrations of pastoralists into agricultural regions usually resulted in sedentarization. Thus it may be surmised that in the third and second millennia B.C. in the Near East there existed herdsman husbandry and semi-nomadic pastoral­ ism, but in no way was there any real nomadism. Purely nomadic groups, if they did exist, were a rare exception (Rowton, 1967:109; Klengel, 1972:177). But despite the opinion of Coon (1976:53) these ancient semi-nomads bore little resemblance to contemporary 'Halbbeduinen' to use the terminology of German scholars. Half-Bedouins have used camels as transport animals, even in cases where they themselves do not breed camels and, moreover, various economic and political relations have always linked them to pure Bedouins. For the time being it cannot be stated categorically that Arabia was the only centre of nomadization in all the areas of the Near East being examined here, but all the evidence supports the suggestion that it was the major centre. In certain respects the role of the Arabian peninsula in the development and dissemination of nomadism in the Near East was similar to that of the Sahara in the development and dissemination of African nomadism. Just as in the Sahara, on the petroglyphs first discovered by the expedition of Ryckmans-Philby (Lippens, 1956), cattle appear in the sixth millennium B . C , small stock appears in the fourth and third millennia B . C . and from the end of the third millennium B.C. definitively replaced cattle in Central Arabia. No later than the fourth millennium B . C pastoralism becomes the predominant economic activity here (Anati, 1968; Anati, 1974; Tchernov, 1974). However, in Arabia these dates also should be regarded for the time being as tentative ones. In the opinion of McClure (1971:3-5,28-9), the climate of Arabia altered at the same time as the climate in the Sahara did, thus it is possible to talk about a single 'Saharo-Arabian belt'. The last dry phase began about 2500 B.C. and continues up to the present day (cf. Butzer and Twidale, 1966:135; Pearse, 1971:14). It is not known how this new dryness affected the ancient pastoralists of Arabia. Perhaps one consequence of it was the movement of pastoralists to the borders of agricultural areas (Oates, 1976:30). One thing is clear. This is that the process which came to be called the 'bedouinization of Arabia', that is the final stage of the formation in Arabia of real pastoral nomadism, was linked to the beginning of the utilization of the camel as a riding animal. The exact time and place of the domestication of the camel (as well as the fust attempts to herd the camel) are both practically unknown. It is also unknown whether the two were connected and how the dromedary

The origins of pastoral

nomadism

(Camelus dromedarius) and the Bactrian camel {Camelus bactrianus) were domesticated. The breeding of camels in Western Asia is usually considered to have begun in the middle or even in the second half of the second millennium B . C . (Albright, 1942:96,100; Bacon, 1954,47-8; von Wissmann, 1959:880b-881a; Coon, 1976:50). In the famous texts from Mari (eighteenth century B.C.) reference is often made to the donkey, but never to the camel. However, reference is made to the latter in the earliest parts of the Bible. Admittedly, these references are sometimes regarded with scepticism; since the nineteenth century they have frequently been declared an anachronism (for one of the most recent of such statements see Klengel, 1972:154; cf. Albright, 1960:207). However, there is insufficient foundation for such scepticism. In the Bible the camel is referred to 56 times, 25 of them in just one book of Genesis; moreover, all these references relate to the time of the patriachs. Camels are included in the stock of Abraham, Isaac and Jacob (see, for example, Genesis 12.16,24.10,24.35,30.43,32.15 and others). It is noteworthy, however, that here camels do not yet figure as riding animals. At the same time commentators come to the conclusion from the report in Genesis 32.15 of Isaac's gift of 'milch camels' that, at the time of the patriarchs, milk was already being used as food. Very conspicuous, however, is the fact that the number of camels kept by the patriarchs (Genesis 24.10 - a herd of which ten is by implication a tidy proportion; 32.15 - thirty) was moderate in comparison to the herds said to have belonged to Job (Job 1.3 - 3,000; 42.12 - 6,000) and particularly to the 50,000 camels taken from the Hagarites by the descendants of the tribe of Reuben (I Chronicles 5.21). Some scholars also surmise that the domesticated camel is to be found in a number of images which originated in various areas of the Near East and are to be dated earlier than the middle of the second millennium B.C. Thus the opinion was voiced some time ago that the time of the domestication of the camel must be put at an earlier date (Free, 1944; Rathjes, 1955:18b, 114-15; Pohl, 1957:165-6; Bulliet, 1975:36), and according to Ripinsky (1957) even as far back as the fourth millennium B.C. Henninger (1969:42) is probably right to consider that two questions should be singled out: the date of the domestication of the camel and the date when camel-breeding as a specialization began. There was evidently a great time lapse between these two events. Be that as it may, it was only specialized camel-herding, which began in Arabia somewhere in the middle of the second millennium B.C. and was possibly stimulated by the desiccation of the peninsula, which led to the dissemination of real pastoral nomadism in the inner regions of Arabia and Syria. Already in the eleventh century B.C. the Midianites, Amalekites and 'children of the East' stormed through Jordan into Palestine (Judges 6.1-6), '. . . for both they and their camels were without number' (Judges 6.5). 100

The origins of pastoral

nomadism

According to Dostal (1959:22) their invasion was linked to the perennial problems faced by nomads: the desire to extend grazing territory for the increasing numbers in their herds, the need of agricultural products and of trade with sedentary societies, etc. As to the horse, there is information that it was present in the cultivated areas of the Near East by the beginning of the second millennium B.C, and perhaps in an even earlier period (Salonen, 1955:17-18; Zeuner, 1963:317; Moorey, 1970); but the horse long remained a precious and rare animal. The question of the role of the Indo-European tribes in its dissemination is still a subject of debate (cf. Kammenhuber, 1968; Diakonoff, 1972; Mayrhofer, 1974). However, at this point it is more important for me to emphasize other circumstances. Firstly, in the second millennium B.C. the horse was almost never used as a riding animal. The second millennium B.C. was the time when military chariots became very widespread. It was only the last Assyrian kings who hurriedly tried to create a cavalry, but they did not manage to do this with any great success (Yadin, 1963:286-7; Khazanov, 1971:56). Secondly, in the Near East in the second millennium B . C and, indeed, later horses, which were herded primarily in mountainous regions (Jankowska, 1969:266-7), were less utilized by mobile pastoralists than in sedentary agricultural states where they were used primarily for military purposes. It is thought that in Arabia the horse only appeared at the beginning of the first century A . D . (Hitti, 1956:20-1). Strabo emphasized that there were no horses in Arabia and that their work was done by camels (Strabo XVI.4.2; XVI.4.26). When was it that pastoral nomadism in Arabia finally assumed that form which became the specific form of the Near Eastern type of nomadism? Dostal (1959:20ff.; cf. Bulliet, 1975:87ff., 113ff.) surmises that at the end of the first millennium B.C. a new saddle was devised for the camel in Northern Arabia; this saddle had a pommel and was placed on the animal's hump leaving the arms of the rider free. As a result the efficiency of camel-warriors was significantly increased and they became a formidable military force. Dostal considers that it is from this period onwards that we can talk about real Arabian Bedouins. Henninger (1969:37) dates 'die voll beduinische Period' from the third century B.C; Coon (1959:872) dates it even later, to the time between Christ and Mohammed. The cited materials prevent us from fully agreeing with Caskel (1954:36-46) who links the bedouinization of Arabia with the destruction of sedentary states and the decline of caravan trade in Arabia resulting from the general crisis in the ancient world which, according to Caskel, incited part of the sedentary population of Arabia to become nomadic. However, it is not impossible that the number of nomads at the end of the ancient period and the beginning of the medieval one could, indeed, have been augmented by sedentary peoples. 101

The origins of pastoral

nomadism

As a whole it may be concluded that the Near Eastern type of nomadism, despite its prolonged prehistory, was finally formed no less than a thousand years later than the Eurasian steppe type, although this happened without any essential influence from the latter. With regard to their genesis there does exist between these two types still one more important difference. Whilst the Eurasian steppe type of nomadism back in antiquity, in the first millennium B.C., basically had already occupied almost all of the available habitat (Khazanov, 1975), the territory of the Near Eastern type was only established in the Middle Ages after a whole series of wars, invasions, migrations and the consequences which all of this entailed. However, in the preconditions to the formation of these types of nomadism a certain parallelism may be observed. Both in the Eurasian steppes and in the Near East the development of pastoral nomadism was out of mixed economy and through extensive herdsman husbandry and/or semi-nomadic pastoralism. In both areas the earliest forms of extensive pastoralism were based in the first instance on the herding of small stock. In both of them the horse and the camel had analogous functions in the economy. They were thoroughly utilized in both regions as food, as well as for transport and riding purposes. However, it was only after a long time that both were fully utilized in both regions. In both regions the immediate transition to pastoral nomadism in its traditional economic forms, at least with regard to species-composition of herds and methods of their utilization, was probably linked to specific changes or variations of climate. Finally, in both regions nomads were linked right from the beginning in a complex system of peaceful and hostile relations with sedentary states. The Middle East (Asia Minor, Iran, Afghanistan). The origins of pastoral nomadism in Asia Minor, Iran and Afghanistan differ very considerably from those of pastoral nomadism in the Near East. On the territory of the latter nomadism emerged independently due to the decisive role played by the Arabian centre. In the Middle East nomadism did not so much emerge as spread as a result of external impulses and, particularly, of direct migration, while at the same time it was influenced by social and political factors. Thus the region in question is an intermediary one between the Eurasian steppe and Near Eastern regions on account both of the economic and cultural forms of nomadism there and of its genesis. Herdsman husbandry and possibly even semi-nomadic pastoralism, particularly in itsyaylag variants, appeared here very early, no later than the third millennium B.C. in the mountainous regions of Iran and on the Armenian plateau. But there are not the facts to substantiate the opinion of those scholars who believe that real nomads appeared here in the third millennium B.C. (see, for example, Spooner, 1972:126; Spooner, 1973:37). It would appear that neither the Gutians nor the Kassites who played an important role in the history of Mesopotamia were nomads. The few written 102

The origins of pastoral

nomadism

sources which refer to mobile pastoralists in the Middle East, even in the first millennium B . C . and the first half of the first millennium A . D . , present a picture which is very different from the Middle Ages. In antiquity there were no real nomads in Asia Minor, although there were pastoralists practising transhumance (de Planhol, 1959:526; de Planhol, 1969:71-2). In Iran, according to Herodotus, four out of six Persian tribes were nomadic (Herodotus 1.125). However, the father of history supplies no details about these tribes. From other sources it would appear that in antiquity there were fewer real nomads (camel-herders in Fars, referred to by Strabo XV.3.1, for example) than extensive pastoralists. The latter first and foremost were mountain dwellers who caused a great deal of trouble for various powers, particularly the Greeks. It was not without reason that they were frequently called 'vagrants' and 'robbers' (see, for example, Polybius V.44; Strabo XI. 12.4; Strabo XI. 13.3; Arrian, Anabasis, III.24.2-3). The way of life of the Mardeans, Cadusians, Cossaeans, Uxians, Elameans, Paraetaceni and others was in some respects similar to that of the Kurds and Lurs later on, and partly even of the Bakhtiari. Basically their pastoralism was the yaylag type and was supplemented by the secondary occupation, agriculture (Briant, 1976; Briant, 1977). Evidence of this is provided by the Cossaeans, Paraetaceni and Elameans (Strabo XV.3.1; Strabo XVI.1.18; Diodorus XVII.111.4; Arrian, Anabasis, VII. 15.2). Arrian writes of the Uxians that'. . . they had no money nor arable land, but they were for the most part herdsmen'; but at the same time he notes that the Uxians live in villages (Arrian, Anabasis, III. 17.3-6). Curtius Rufus describes how the Mardeans (one of the nomadic Persian tribes, according to Herodotus) dug caves in the mountains and hid in them with their wives and children (Curtius Rufus V.6.17); but he tried to emphasize their savagery. According to Diodorus (FHG, 90, F66) and Strabo (XV.3.1) some of the Mardeans practised agriculture. In the species-composition of the herds of these mountain tribes small stock played an important role alongside large stock. However, there were few riding animals, few horses and very few camels. In written sources the mountain tribes usually appear as unmounted archers (Strabo XVI.1.18; Xenophon, Anabasis, IV.3.28-33), and only the Cadusians are described as possessing a cavalry (Plutarch, De Alexandri Fortuna, 1.329b). The tribute Alexander demanded from the Uxians was 100 horses, 500 transport animals (oxen?) and 30,000 sheep (Arrian, Anabasis, III. 17.6). Real nomads of the steppe type appeared in Asia Minor, Iran and Afghanistan from outside, from the Eurasian steppes. However, the 6

* It is curious thiit the habit of using caves as winter shelters for livestock has been preserved up to the present day by semi nomadic Kurds and Lurs (de Planhol, 1966a:296; C o o n , 1976:215; Hole, WH:I52).

103

The origins of pastoral

nomadism

Cimmerians, Scythians, Parni, Sakas, Kushans, Alans, Chionites, Ephthalites and others exercised no essential influence on local forms of pastoralism, because when they did settle in new territories they frequently became sedentary like, for example, the Sakas in Seistan. The position changed in the Middle Ages when the invasions of great masses of nomads from Arabia and the Eurasian steppes and the social and political upheavals connected with them (particularly the disintegration of the power of agricultural-urban societies and the growth of different nomadic polities) led to the territorial expansion of nomadism and an increase in the number of nomads in this region. De Planhol (1969:73) writes of'la bedouinisation medievale' of Anatolia and Iran. Only if Bedouins are understood to mean all pure nomads, which is hardly appropriate, is this notion tenable. Thus it is more apposite to refer to the medieval nomadization of Asia Minor, Iran and also Afghanistan, for the Arabian Bedouins played a less significant role in this process than did the Turks and Mongols of the Eurasian steppes (Petrushevsky, 1960:41ff.). The nomadization occurred in different ways. The principal one was by way of the direct migration of nomads into the Middle East and their appropriation of land there for nomadic forms of pastoralism; this was frequently accompanied by the pushing out of the sedentary and semi-sedentary peoples from the areas which the nomads had selected. Marco Polo (Ch. XXXV) noted when describing his journey between Kirman and Hurmuz: 'In former days there were plenty of inhabited places on the road, but now there are none; and you meet with only a few people looking after their cattle at pasture' (The Book of Ser Marco Polo, 1975:91). During their migrations in the Middle Ages the Baluch frequently wreaked destruction on the local population of agriculturalists (Pikulin, 1953:23). The nomads strove to preserve their economy in the forms to which they were accustomed, but ecological constraints sometimes prevented this. The Bedouins could not penetrate the Anatolian plateau for, as it is testified by al-Jahiz, it was too cold for their dromedaries. In Iran they confined themselves to occupying the southwest province of Fars (according to al-Istakhri there were some 500,000 nomad kibitkas in Fars in the tenth century), which from antiquity had bordered onto lands used by camel-herders (Strabo XV.3.1), and partly also Khuzistan (de Planhol, 1969:73; Cahen, 1975:310). The majority of Bedouins who were allotted lands in Khurasan (in about 672-3 some 200,000 Arabs were settled there Zarrinkub, 1975:28) gradually sedentarized and became Iranicized. But the Turks with their Bactrians which were more accustomed to the cold and, the more so, with their horses were able to settle themselves in Anatolia, particularly after the Saljuq conquest (de Planhol, 1966a:302). It is noteworthy that the majority of toponyms of Oghuz origin are concentrated in Central Anatolia, where the plains suited nomads. Only 104

The origins of pastoral

nomadism

after they had occupied these plains did the Oghuz move further on through the Western Anatolian passes to the Aegean Sea, and only in the thirteenth century did they move through the Lycia and Cilicia mountains and reach the valleys of the Mediterranean (Eremeev, 1971:86). However, in Anatolia, particularly Western Anatolia, where natural conditions favoured agriculture and the policy of the central powers, irrespective of their origins, almost always was to encourage sedentarization, pastoral nomadism was significantly less widespread than it was in more arid Iran (de Planhol, 1959:525-9). Nomads moved into Iran over the course of many centuries, particularly large masses of them arriving in the eleventh and twelfth centuries with the Saljuqs (in Northern Khurasan, Gurgan, Dihistan, Azerbaijan, Arran, certain areas of Kurdistan and Luristan - see Lambton, 1953:59; Bosworth, 1968:79) and later, in the thirteenth century, with the Mongols. Between the twelfth and fourteenth centuries nomads made up about one quarter of the population of the country (Helfgott, 1977:36). In addition, the central powers often made them re-group and settle in new places, thus enabling nomadism to spread further (see, for example, Lambton, 1953:131). The privileged position occupied by nomads in Iran and the damage that they brought to the agricultural population there led to the nomadization of separate groups of the local sedentary and semi-sedentary population. Thus the greater part of the Lurs and Bakhtiari was formed of people who previously had been agriculturalists practising herdsman husbandry and who had had to leave their villages in the Middle Ages (see de Planhol, 1966a:292; Trubetskoi, 1966:175). It is possible that there were separate nomadic or semi-nomadic groups in Southern Afghanistan before the tenth century. However, it was only after the Oghuz migration and, particularly, the ensuing invasions of the nomads of the Eurasian steppes (we only need remember the Nigudaris who settled in Eastern Khurasan) who infiltrated the local population and affected their pastoralism, that nomadism in Afghanistan assumed its traditional forms (Schurmann, 1962:45-6). Some nomads were crowded by others. Thus the Saljuqs pushed the Baluch out of Khurasan and Kirman into Makran and further east (Pikulin, 1953:18; Frye, 1961:47). In this way the chain reaction of nomadization, which took place in the Eurasian steppes in the first millennium B.C, was in the Middle East a protracted and interrupted process which was completed two thousand years later. Only in the second millennium A.D. did nomads here occupy the whole ecological zone which suited them. It was only in the Middle Ages that the three basic variants of nomadism and semi-nomadism in the Middle East finally were formed. The first variant basically consists of the herding of sheep and horses and pastoral migrations ol the meridional t y p e in the steppe and semi-desert zone (Gurgan, Mughan 105

The origins of pastoral

nomadism

and Khurasan), and was most influenced by the Eurasian steppe type of nomadism. Back in the period of their conquest of Iran the Mongols decided to use the Mughan steppe as their winter quarters because there was good fodder for their horses there (Galstian, 1977:167, 170, 176). The second variant is connected with camel-herding and the cultivation of date palms (the Arabs of Southwest Iran, the pastoralists of Makran and Baluchistan), and was influenced by the Near Eastern type of nomadism. In Baluch tribes, particularly those like the Rind, there are groups of Arab origin who were incorporated during the period of the seizure by the Baluch of the Makran coastline (Pikulin, 1959:17). Already in the tenth century Baluchistan was famous for its camels which were even exported to Khurasan and Fars (Barthold, 1971:92). The third variant is the mountain one (the Lurs, Bakhtiari, Qashghai, Kurds and others), which more than other variants is founded on the herding of small stock, and in which many old local characteristics are preserved (de Planhol, 1966a:298), although here large stock was more widespread in ancient times than it is in the modern period. As a whole the nomads of the Eurasian steppes exercised a greater influence on the nomadism of Iran and Afghanistan than the Bedouins of Arabia. The fact that many pastoral terms in the Iranian language are borrowed from Turkic languages is, therefore, no coincidence. Evidently such an influence was determined both by ecology and by history. However, it was in the Middle Ages that the dromedary began to predominate over the Bactrian camel in the Middle East, while in the first millennium B.C. the camel with two humps was known in almost all of Iran (Bulliet, 1975:157). Africa. Although there are different types of pastoral nomadism in Africa, the genesis of these is in many ways interconnected; thus it makes sense to examine them in one section. It is now evident that the sources of pastoralism within the continent itself are to be found in Egypt and North Africa. It is not clear, admittedly, to what extent they are linked to local development, and to what extent they are linked to impulses or even direct migrations from Asia (cf. Mauny, 1967:583-99; Clark, 1970:197). However, in ancient Egypt pastoralism always remained an adjunct of agriculture. The centre of the development of extensive forms of pastoralism in North Africa was in the Sahara. It is not impossible that domesticated animals appeared in North Africa at a fairly early date, in the seventh millennium B.C. (Higgs, 1967:167); and in the fifth and fourth millennia B.C. the inhabitants of the Sahara and Sahel even (Smith, 1978:94) were pastoralists who herded long-horned and short-horned breeds of cattle alongside sheep and goats (Clark, 1970:198; Clark, 1972:132; Lhote, 1976:67). The famous Saharan rock paintings have popularized the idea that already in the Neolithic period the Sahara was inhabited by pure nomads who were cattle-keepers. 106

The origins of pastoral

nomadism

I doubt that this can be true. There is no certainty that the rock paintings present a correct picture of the species-composition of herds; even less are they able to prove the absence of agriculture. There is certain archaeological evidence to support this supposition (Clark, 1967:605; Hugot, 1968:488). Admittedly, it is possible that at this time pastoralists in the Sahara, especially individual groups of them, were fairly mobile. However, the fact that they herded no riding animals like the horse and camel meant that their mobility was limited. The time at which the donkey appeared in North Africa has not been precisely established (Nicolaisen, 1963:100, 484). Moreover, semi-nomads rather than pure nomads tended to herd donkeys. The following stage is linked to the desiccation of the Sahara which, evidently, increased in the middle and second half of the third millennium B . C . (cf. Lamb, 1966:6). Since the first millennium B . C . natural conditions in the Sahara have changed little, if at all (Butzer and Twidale, 1966:135-7; Butzer, 1966:78; Shaw, 1971:59; Kellogg and Schreider, 1977:147). As the Sahara became drier the population there was forced to move to the South and East. The population which remained had to adapt to the changed conditions; this entailed an increase in the role of pastoralism in the economy and in the degree of mobility of pastoralists. Can it be assumed that in the third and second millennia B.C. the Sahara was already inhabited by real nomads? There is too little data to say either way. It is possible that there were groups which had completely broken with agriculture, but we can scarcely talk about the formation in this period of purely nomadic societies. According to the traditional view, the horse appeared in Egypt about 1700 B.C. and from there rapidly spread into other countries of North Africa (Hancar, 1955:483-4); it is possible that we should put this date back a few centuries (Emery, 1960: 8-9). But in the second millennium B.C. the domesticated horse everywhere was used mainly for chariotry, not for riding; in North Africa chariots were associated with the still mysterious Garamantes. In North Africa riding supplanted chariots only in the middle of the third century B.C. (von Wissmann, 1959:889). The last time Libyan chariots were used was in 307 B.C. (Diodorus XX.64.3), and the Libyan cavalry is mentioned for the first time in 261 B.C. (Polybius 1.19.2-4). On account of the specifics of natural conditions the horse played a lesser role in the opening up of the Sahara by nomads than the camel did later on (Briggs, 1960:21; Nicolaisen, 1963:112-13). Be that as it may, already in the third and second millennia B.C. pastoralism in North Africa differed essentially from later East African pastoralism because cattle were not the most important species in the former. For the time being the emergence of real pastoral nomadism can most circumspectly be dated to the beginning of the first millennium B.C. It was at this time that African nomads first came to the specific notice of the authors 107

The origins of pastoral

nomadism

of antiquity who outrightly refer to part of the Libyan tribes as nomads. Judging from the data of Hecataeus (FHG, F335) and Herodotus (IV.186, 187,191) in the eastern part of North Africa there lived nomads, and in the western part there were agriculturalists. The former the Greeks called kreophagoi, that is meat eaters, and galaktophagoi, meaning milk eaters, while the latter they called sitophagoi, meaning grain eaters. Nevertheless, only with the introduction of the camel into the pastoral economy did the nomadism of North Africa assume its finished form. This process took place over a fairly extended period. It is thought that the camel appeared in Egypt in the sixth and fifth centuries B.C. or somewhat earlier, during the Assyrian conquest (Hitti, 1956:22; Nicolaisen, 1963:314; Coon, 1976:25), and rather later in the other countries in North Africa. Hitti, for example, links this event only with the Muslim invasion of the sixth century A.D. But, as it has already been noted, judging from the Bible and certain archaeological data, camels could have appeared in Egypt in the time of the patriarchs. It seems that the dromedary is first referred to in North Africa by Julius Caesar in connection with the Battle of Thapsus. Some scholars date the spread of the dromedary over North Africa to the first centuries A.D. (Capot-Rey, 1953:85; Nicolaisen, 1963:314). But before the arrival of the Romans in North Africa camels were evidently already known there, having been brought in from Egypt and the Sahara (Demougeot, 1960:241-7). It is likely that for some relatively short period of time, for a maximum of a few centuries, the camel co-existed with the horse in North Africa (Monod, 1932:96, 99; Mauny, 1954:14). However, the future was the camel's. According to Julien (1956:160; cf. Levtzion, 1971:120-1), already in the first centuries A.D. in North Africa camel-herding tribes were forming in place of former sheep-herding ones and were being forced by the pressure of the Romans into the Sahara. It is from this period that it is possible to talk about the principal similarities between North African, particularly the Saharan and Arabian, pastoral nomadism. In the Middle Ages the two became even closer. In the fifth or sixth centuries A . D . the Saharan nomadic camel-herders pushed into the zone which the Romans up to then had regarded not only as their own territory, but also as an area of agricultural civilization (Capot-Rey, 1953:186-7). The Arab conquest of North Africa in the seventh century added little to this. However, the subsequent appearance here of Bedouin tribes, particularly the Hilalian migration of the eleventh century, and the ensuing consequences (de Planhol, 1968:140ff.) signified the completion of the long process. The dissemination of pastoralism and food-producing economy in different areas in Africa generally occurred, it is now clear, considerably earlier than until not long ago it was thought to have occurred. However, the 108

The origins of pastoral

nomadism

establishment of pastoral nomadism in these areas was also a very protracted process. Up to now there has been insufficient data to establish with any certainty the time and conditions in which pastoralism emerged in Nubia. At the moment the general view is that pastoralism and agriculture appeared here in the fifth and fourth millennia B.C. as a result of the influence of Egypt (Trigger, 1976:31-5; Krzyzaniak, 1976:762). This does not eliminate the possibility that the food-producing economy in Nubia was not later influenced by other external impulses linked to the movement of people from the Sahara which was becoming more and more dry (Arkell, 1961:42-52; Clark, 1967:613; Clark, 1970:206) and from the Arabian peninsula. It is possible that pastoralism became the basic occupation of the population of North Nubia as far back as the fourth and third millennia B.C. (Arkell, 1961:32,48-53). In Egyptian inscriptions of the sixth dynasty there are references to pastoral tribes which sometimes are looked upon as the direct descendants of the ancient Blemmyes and the ancestors of the Beja to today (Paul, 1954:21, 27ff.). But once again I should like to emphasise that there are insufficient grounds for us to look upon these ancient pastoralists as real nomads. The horse appeared in Nubia at the end of the Middle Kingdom (Emery, 1960:8-9), but evidently it was not used as a riding animal and did not play an important role in the pastoral economy. Pastoral nomadism can only be connected with any certainty with the territory of what is now Sudan from the first millennium B . C ; at that time Strabo describes 'Ethiopian tribes' and wrote that some of them were nomadic. 'And indeed the Aethiopians lead for the most part a nomadic and resourceless life, on account of the barrenness of the country and of the unseasonableness of its climate and its remoteness' (Strabo XVII. 1.3; see also XVII.2.1; cf. Diodorus III.8). The camel, however, appeared in Nubia only at the end of the first millennium B.C. (Arkell, 1961:163). Only after that is it possible to talk of the appearance of separate traits of Near Eastern nomadism in the Sudan. At the same time it may be surmised that already in ancient times there were features in the pastoralism in this region which were later to characterize the East African type of nomadism. Diodorus Siculus wrote of ikhorphagoi (blood eaters) and stated that, '. . . they live on milk boiled with blood' (Diodorus III.15). In this way already in ancient times the Sudan was a marginal zone in which different types of nomadism bordered onto and influenced each other. In the Middle Ages this state of affairs developed further. However, this cannot be explained by ecology alone, for in part at least it. was also linked to purely historical factors - external impulses, migration, etc. 109

The origins of pastoral

nomadism

No less complex, if not more so, were the genesis of pastoral nomadism and its subsequent history on the territory of the Horn of Africa. Clark (1967:613, 615; 1972:137) surmises that domestic animals and plants appeared on the territory of what is now Ethiopia and Somalia from the Nile valley at least in the second millennium B.C., if not in the third. It is likely that the food-producing economies of the Horn gradually moved away from a complex agricultural-pastoral or pastoral-agricultural economy towards a pastoralism in an increasingly extensive form. The specificity of this process was basically determined by three factors: the increasing desiccation of the Horn (Clark, 1954:149-50; Lewis, 1955:71), the present-day natural conditions in the Horn, for in East Africa only Somalia has a really arid climate (Butzer, 1966:80), and, finally, its geographical and historical proximity to Arabia. At present there is insufficient archaeological data for us to trace the early stages of the development of pastoralism on the territory of the Horn of Africa. Rock paintings can only fill part of this gap. The earliest rock paintings in Eritrea, Harar and Somalia depict scenes with herdsmen. They depict long-horned, humpless cattle. It is only considerably later, possibly in the first millennium B . C , that the zebu appears in the paintings, and even later, in the last centuries of the first millennium B . C and the first centuries of the first millennium A . D . , that the camel appears. Clark (1967:615; see also Cole, 1963:267-8; Clark, 1954:315; Graziosi, 1964:91-8,187-90) links them to the appearance of migrants from Arabia. It is likely that in the Horn of Africa pastoral nomadism developed in a direction which was close to the Arabian sub-type of nomadism and more detached from the East African type. In Chinese sources of the ninth century it is written of the inhabitants of the Horn that,'. . . they prick the vein of one of their oxen, mix the blood with milk and eat it raw' (Freeman-Grenville, 1962:8). This custom has been preserved to the present day amongst the Galla and the Southern Somali. But amongst the Northern Somali, who are pure nomads, it does not exist (H. S. Lewis, 1966:30 n . l l ) . Ibn Battuta in the fourteenth century recalls nomads pasturing camels and sheep living in the Horn (Freeman-Grenville, 1962:27). It is evident that those features of pastoral nomadism in the Horn of Africa which enable us to relate it to the Near Eastern type of nomadism were finally formed only in the second millennium B . C , with the settling of the Somali in new territories (Lewis, 1955:45-6; Lewis, 1960; Lewis, 1961: 24-5; cf. H.S. Lewis, 1966). The diverse influence of Arabia on the formation of the specific characteristics of pastoral nomadism in the Horn also could have contained one ideological feature - the dissemination of Islam. The formation of the East African type of pastoral nomadism was also, it 110

The origins of pastoral

nomadism

is now clear, considerably more protracted than not long ago it was supposed to have been. Judging from the most recent data food-producing economy already began to spread south and southeastwards from Nubia in the second half of the second and in the first millennium B . C . In the first millennium B . C . in Kenya and the Rift Valley area there were pastoralists who, to a limited extent, also cultivated millet (Clark, 1967:615, 617; Clark, 1970:208; Sendon, 1968:491; Sutton, 1971:144; Jacobs, 1975:410; Phillipson, 1976:66f.; Oliver, 1977:653), and towards the beginning of the first century A . D . food-producing economy had even reached South Africa (Juskeep, 1971:249; cf., however, Troup, 1975:15). Linguistic data also testifies to the fact that pastoralism was known in the southern part of East Africa and certain areas of South Africa before the Bantu appeared (Ehret, 1967). In Zimbabwe pastoralism played an important role, but it was not real pastoral nomadism. As a whole the economy was a mixed one, although it is possible that there were groups which specialized in pastoralism (Garlake, 1978). In this way the migrations of pastoral peoples in East Africa at the end of the first and the beginning of the second millennium A.D. were one stage in a protracted process, rather than the beginning of the process (Posnansky, 1967) which continued to the nineteenth century. The Maasai only appeared in East Africa in the seventeenth century (Oliver, 1977:654), and they also were not the last to do so. Somewhere on the border of the first and second millennia A.D. in the uplands of Ethiopia or in wider territories, including Southern Sudan, a new breed, the long-horned, humpbacked sanga, was bred by crossing long-horned cattle without humps with the short-horned, humpbacked zebu. Along with pastoralists migrating from the north the sanga arrived in Central and East Africa and here became the most common breed of cattle (Posnansky, 1966:88-9). However, the formation of the nomadism of the East African type was completed not long ago at all, in the second half of the second millennium A . D . The reasons for these pastoral migrations into East Africa at the end of the first and the beginning of the second millennium A.D. , as is the case with the later ones, are not very clear. Oliver (1961:59) offers the usual explanation of increase in the number of livestock in herds and insufficient pastures. It is possible that the dry climate in certain periods should also be taken into account (Jacobs, 1975:411). According to Dale (1954:26-8), in East Africa the climate began to get wetter and forests began to spread only in the fifteenth century. If this is the case, then the formation of the East African type of nomadism is linked, at any rate indirectly, to climatic changes. Northern Eurasia. I have already noted that the view that reindeerherding is the most ancient form of pastoralism is a mistaken one. Certain 111

The origins of pastoral

nomadism

scholars, admittedly, still insist that in the Far North reindeer-herding emerged independently and some of them further surmise that this occurred at a very early date (Krader, 1970:321; Vdovin, 1973:242-5; Dikov, 1975:52; Simchenko, 1976:77-8). However, the fact is that this opinion cannot be seriously argued. I give preference to the argument that reindeer-herding emerged in South Siberia amongst peoples who had the skills for a pastoral economy (see, for example, Vasilevich, 1969:78-80; Vainshtein, 1980:120). If the Bolshaia Boiarskaia pisanitsa really do depict reindeer, then the domestication of reindeer took place no later than the first millennium B.C. But in the opinion of Griaznov (1978:224), who disputes the view of such well known zoologists as Vereshchagin, Flerov and Tsalkin, when the deer's head is held high on a long neck this is a sign of red deer (Cervus elaphus, cerf in French, Hirsch in German), and not of a reindeer (renne in French, Renntier in German). In Chinese sources there are references to deer being herded in the fifth century A . D . in Fusang (Bichurin, 1950a:47). Flor (1930:133) gave some thought to the domestication of the deer in the Sayan Mountains, but he traced the origins of reindeer-herding back to hunting. The Sayan and Altai mountains, in which reindeer-herding continues up to the present day and in which there lived in ancient times the ancestors of the Samoyed (Samodian) peoples, were if not the only then the oldest centre of reindeer-herding (Vainshtein, 1980:120ff.). However, reindeer-herding was not the basis of the economy there. It was only when the Samoyeds were forced by the pressure of Turkic-speaking peoples to move northwards, into the taiga, and mixed with the aboriginal population that the practice of herding reindeer became widespread in the western and central parts of the taiga zone. It is thought that this happened at the turn of the first and second millennia A.D. when the only domestic animal the Samoyeds continued to keep was the reindeer which could adapt to its new environment. Nevertheless, the herding of large numbers of reindeer was impossible in the taiga. Thus the Samoyeds were forced to transfer to food-extracting forms of economy, hunting and fishing, and reindeer were mainly used as transport animals (Maksimov, 1929:22ff.; Vasilevich and Levin, 1951:78-80; Khomich, 1966:39; Vainshtein, 1972:108-9). However, it is not impossible that the spread of reindeerherding in North Eurasia began considerably earlier. Simonsen (1967:70) suggests with some circumstantial data that the Lapps were herding reindeer back in the Kjelm0y period (200-600 A . D . ) . Judging from archaeological data it is not impossible that the people living around the lower reaches of the Ob at the turn of the eras B . C and A . D . hunted, using domesticated reindeer as decoys (Moshinskaia, 1953:78ff.). 1

7

The word pisanitsa rock.

112

is a fixed Russian term meaning a collection of petroglyphs on a single

The origins of pastoral

nomadism

Some peoples in the western and central parts of the taiga and tundra zene copied the Samoyed practice of herding domesticated reindeer, and through the Tungus the peoples of the tundra zone of Northeast Siberia also learned this practice. Admittedly, it is still not clear whether reindeer-herding appeared amongst the Tungus independently, although influenced by Turkic peoples who herded horses, or whether it was the result of borrowing from the Samoyeds (cf. Vasilevich and Levin, 1951:84-6; Vasilevich, 1969: 80; Vainshtein, 1972:114-22). Be that as it may, by the sixteenth and seventeenth centuries the Nentsy, Komi-Zyrians (Izhmans), Khants and Mansi, the ancestors of the Entsy and Nganasans, Kets, Yukagirs, Koriaks and Chukchi already knew about reindeer-herding. However, at this time all these peoples herded reindeer primarily so they could use them for transport; variants of food-extracting hunting and fishing economy prevailed (Vasilevich and Levin, 1951:82-4; Levin, 1958:221-2; Istorikoetnograficheskii atlas Sibiri, 1961:24ff.; Alekseenko, 1967:69-70; Obshchestvennyi stroi u narodov Severnoi Sibiri, 1970:39; Shnirelman, 1974:50-1; Khomich, 1976:146; Krupnik, 1976a:57-65). Thus, although the inhabitants of the taiga and tundra zones of Eurasia were familiar with domestic animals and kept them in limited numbers, over the course of many centuries their economy continued to be a basically food-extracting one. A sudden break with this tradition began at the beginning of the eighteenth century. Some interesting works by Krupnik (1975; 1976; cf. Leeds, 1965:98), with whose conclusions I agree in principle, are devoted to describing and explaining this break. From the beginning of the eighteenth century in almost all of the tundra zone there was an abrupt increase in the number of domestic reindeer. Consequently, the transition from different forms of hunting and fishing economy, with supplementary herding of small numbers of reindeer for transport purposes, to reindeer-hunting and the herding of large numbers of reindeer for transport purposes occurred. Finally, at the turn of the eighteenth and nineteenth centuries came the transition to nomadic reindeer-herding. This entire process was completed in some 100-150 years. Moreover, these changes took place against an evidently favourable background, namely the strengthening of Russian power in the Far North which entailed the establishment of a fixed law and order, the curtailment of inter-ethnic conflicts, the development of exchanges and trade, etc. (Leeds, 1965:124-5). It is possible that excessive hunting was instrumental in the decrease in the number of wild reindeer amongst the Nentsy (Khomich, 1976:75). However, the temperature-fall in the second half of the second millennium A.D. was evidently a direct incentive for the transition to nomadic reindeer-herding (Willet, 1953:55; Brooks, 1954:157; Lamb, 1966:10; Cermak, 1971:17). Some scholars even call this period the Little Ice Age. I 13

The origins of pastoral

nomadism

The fact is that the tundra reindeer, in contrast to the taiga reindeer, is well able to cope with low winter temperatures, but very bad at coping with even very temperate summer ones. Even in temperatures of +10-15° C reindeer stop pasturing, quickly lose weight and become weak. Thus, the temperature-fall created an ecological situation favourable to the develop­ ment of the nomadic herding of large numbers of reindeer. The transition to this new economic system at first led to a sharp increase in the population. Amongst the Nentsy and the Chukchi the population increased fourfold in 80-120 years and then became stable. Discounting the Lapps, nomadic reindeer-herding at the end of the nineteenth century and the beginning of the twentieth was most widespread amongst certain groups of the Komi-Izhmans, European and West Siberian Nentsy, some Chukchi and Koriaks, between ten and twelve thousand people in all (Krupnik, 1976: 58-9). At the same time, as it was pointed out in the first chapter (see p. 42), between the Yenisei and Kolyma rivers there basically continued to live people who were hunters and for whom the herding of reindeer was only a secondary economic activity. The question then arises of why reindeer-herding did not come to occupy the predominant position in an ecological zone which was basically favourable to it. It is possible that the explanation partly lies with a series of geographical and ecological factors; but, first and foremost, it lies with the late emergence of nomadic reindeer-herding which coincided with an increase of pressure and influence from more southern sedentary societies. It is significant that at the end of the nineteenth century and the beginning of the twentieth the herding of reindeer in large herds continued to spread amongst the tundra Entsy, Nganasans and northern Yakuts (Popov, 1948:55; Khomich, 1966:51; Dolgikh, 1970:133-4; Gurvich, 1977:51). The economic rebuilding of the North after the revolution interrupted this process. The time and circumstances of the emergence of reindeer herding amongst the Lapps have not yet been firmly established. Some scholars insist that it emerged independently, others that it was linked to some influence from the East (cf. Vasilevich and Levin, 1951:81-2; Nelleman, 1961. Vorren and Manker, 1962:13, 75; Simonsen, 1972:190, 191; Vainshtein, 1972:124-5). However, amongst some groups of Lapps in Scandinavia the transition to pastoral nomadism began in the sixteenth and seventeenth centuries and even continued into the nineteenth century. An important factor in this process was pressure of agricultural societies from the south, one consequence of which was a decrease in numbers of wild reindeer (Vorren Manker, 1962:57-8, 76-7, 81; Vorren, 1973:185-93). At the same time the products of reindeer-herding were important to the Lapps for trading with Scandinavians (Simonsen, 1967:76). But it is possible that climatic changes favoured economic changes. 114

The origins of pastoral

nomadism

High Inner Asia. At present there is too little data for any firm conclusion to be reached about when and how pastoral nomadism emerged in this area, and about when the specific type of nomadism entailing yak-herding was formed here. We do not really know if these events took place contemporaneously or at different times. According to Hermanns (1949:115, 118, 158, 275, 281) the nomads of Tibet are representatives of 'einer alten, urtumlichen Hirtenkultur, die jetzt in Innerasien gesondert dasteht', who came to Tibet from outside and brought domestic sheep and yaks with them. All that took place in Tibet was the further specialization of the pastoral nomadism. He suggests that the first people in Tibet were nomadic and that only later on did some of them become agriculturalists. Many parts of Hermanns' conception are based not so much on facts as on the theoretical views of the Viennese Anthropo­ logical School, although it is really rather unlikely that the origins of Tibetan nomadism were autochthonous. Nevertheless, there is no conclusive proof that the yak was domesti­ cated in Inner Asia. What is more likely is that its domestication took place in Tibet where to this day wild yaks freely interbreed with domestic ones (Ekvall, 1968:12). All that can be said with any certainty is that the yak was domesticated no later than the first millennium B.C. Yak bones have been uncovered in Pazyryk kurgans in the Altai which are dated to the first half of the first millennium B.C. (Rudenko, 1953:70, 73). It seems that the yak is first mentioned in Europe in the writing of the third century B.C. author, Aelianus, who calls yaks poephagoi, meaning grass eaters. On the other hand, there are grounds for agreeing with Lattimore (1967:211-12) that in Tibet nomadism (I would rather say nomadism in Inner Asia as a whole) did not develop sui generis. Evidently it was formed out of two basic sources: the local mixed economy which possibly had certain pastoral tendencies, and nomadism of the Eurasian steppe type. The influence of the latter was not only in the form of different impulses, but also of direct migrations. From time to time steppe nomads sought refuge in the mountains of Tibet, mainly in the northeast of the country, in Amdo and Northern Kam (Kussmaul, 1962:230ff.; Roerich, 1974:238). Unfortu­ nately we do not know exactly how early this began, but it was evidently back in the first millennium B . C . Lattimore (1967:215) and Ekvall (1968:11) suggest that the Chinese term Ch'iang, which in the first millennium B.C. designated the earliest Tibetan tribes on the borders of China, means 'shepherd'. They believe that these tribes were not yet real high-mountain nomads and that they had not domesticated the yak, although they did herd sheep and cattle. However, M. V, Kruikov has kindly informed me that the symbol of the sheep in corresponding hieroglyphs can, in principle, also be dated back to a totem. 115

The origins of pastoral

nomadism

At the same time there are indications in Chinese sources that the Ch'iang already knew the horse. Part of the Ch'iang was evidently involved in pastoralism of the steppe type until they were pushed deep into the mountains of Tibet. But there were settlements in these mountains when the Ch'iang arrived there (Kychanov and Savitsky, 1975:22-3). Thus, the domestication of the yak is not necessarily to be associated with the Ch'iang. Neither is it quite clear when the Kirghiz emerged in the Pamirs and when yak-herding began to spread there. All that is clear is that this took place in the Middle Ages (Maanaev, 1963:9-11). In his memoirs Babur writes: 'One of the tribes of the wilds of Andijan is the Jlgrak, a numerous people of five or six thousand households, dwelling in the mountains between Kashgar and Farghana. They have many horses and sheep and also numbers of yaks [qiitas], these hill-people keeping yaks instead of common cattle' (The Bdbur-nama, 1922, vol. n:55). It is probable that the ancestors of the Kirghiz borrowed the yak from Tibetans. In Tibetan documents there are references to Kirghiz tribes in regions near Tibet in the eighth century and the beginning of the ninth. In the Middle Ages the amount of territory that Kirghiz tribes lived in was more extensive than it is today; then it comprised areas bordering onto Tibet and Kashmir and included mountainous regions bordering onto Khotan. Even today, in the Kunlun mountains, relatively close to Tibet there are small groups of Kirghiz (Abramzon, 1971:57). Finally, the opinion of a zoologist is apposite here: 'The spread of the yak in Northern Kirghizia, that is the Tien Shan mountains undoubtedly took place from south to north. This leads us to think that the domestic yak in the Tien Shan was more likely to have originated in Tibet than in Mongolia, the more so because south of Kirghizia, in the mountains in Kashgaria and the Kok-shaal-Tau Mountains, there are Kirghiz who herd yaks. Similar, it would seem, were the origins of the yaks herded by the Kirghiz in the Alai and Pamirs . . . Morphologically the yaks in the Alai are in no way different from those in the Tien Shan, and both in turn are closer to the Tibetan yak than they are to the Mongolian one' (Lus, 1930:153). In all the different aspects of nomadism, including its origins, discrepan­ cies are manifested. In the first chapter I looked at the different types of pastoral nomadism statically, in their already completed forms, and defined them as the result of specialization and adaptation to specific natural environments and having limited scope for innovations. Now that the genesis of pastoral nomadism has been examined certain reservations about this definition must be expressed and certain explanations given. The adaptive passivity of pastoral nomadism is relative. Innovations have occurred (for example, the spread of camels in the Middle Ages and the displacement of Bactrians by dromedaries and their hybrids), although 116

The origins of pastoral

nomadism

these have been relatively small and only in part have been for economic reasons (see Chapter 4). Changes in already formed types of pastoral nomadism in the direction of further adaptive specialization, in order to adjust to environment, can be characterized as involution in the sense in which Service (1971:97) employs this term. In addition, elements of active adaptation to changing climatic conditions in the environment are more characteristic of the emergence of pastoral nomadism than they are of the history of its perpetuation. The history of its emergence may be analysed into three components: economictechnological preconditions; the trigger, i.e. specific motivating stimulus for the transition to nomadism; and social and political background. By economic-technological preconditions I mean those changes in a pastoral economy which render it potentially capable of and ready for a transition to pastoral nomadism. Depending on specific circumstances these preconditions may or may not be realized. There has been no determinism about the emergence of pastoral nomadism. Where it has emerged for internal reasons, this has not been the result of any conformity with a set of laws (for example, in the development of a food-producing economy or connected with an increase in productive efficiency to the cost of specialization); on the contrary, the final break with other forms of food-producing economy was an enforced one. Involution was a consequence rather than a cause of the emergence of pastoral nomadism. It is for this reason that I consider that the actual emergence of pastoral nomadism needed a specific motivating stimulus. Evidently, in the majority of cases it was connected with climatic changes. Thus, as opposed to Lattimore and his followers, I consider that nomadism emerged first as the result of adaptation to a natural, rather than to a cultural (social) environment. Nevertheless, if the latter was favourable, the emergence of nomadism was made easier, that is it was possible to overcome the economic and social impediments (on these see Chapter 3) inherent in nomadism. However, this is only one of the ways in which nomadism emerged, and basically it can be regarded as non-diffusionist. It would seem that this was the way the Eurasian steppe, Near Eastern and North Eurasian types of nomadism, independently of each other, were formed. Of course, independently here means only the transition itself to pastoral nomadism, it docs not exclude possible borrowings of domestic animals and correspond­ ing technological skills (that is of things I include in the economictechnological preconditions of nomadism). The second way in which nomadism emerged can be characterized as diffusionist, although its history is also fairly complex and multi-faceted. It comprises the following basic stages: the incipient transition to nomadism in limited local centres in regions other than those being considered; the 117

The origins of pastoral

nomadism

further dissemination of this nomadism in favourable ecological zones (sometimes to the cost of other kinds and forms of economy which are pushed out); the chain reaction, whereby nomadism spreads because of the military superiority of those involved in it or for other reasons; the ability of pastoral nomadism to adapt to new ecological zones with corresponding changes in the species-composition of herds and yet to preserve the previous direction of its economy. Thus, this second way in which pastoral nomadism emerged does not exclude independent adaptive specialization, although the latter is a secondary rather than a fundamental factor in these circumstances. The Middle Eastern and High Inner Asian types of nomadism basically can be regarded as outcomes of a 'diffusionist' process. The position with the East African type is less clear. Moreover, it must also be pointed out that where nomadism emerged independently it first appeared in separate local centres and only then did it spread through the whole of the corresponding region. Thus, in practice, these two ways in which pastoral nomadism developed have many common features. In conclusion I should like to point to one biological aspect of the problem. Multispecialized types of pastoral nomadism, in which animals with high morphological variability were predominant in the speciescomposition of the herds, had the greatest capacity for lability, transforma­ tion and 'expansion'. On the other hand, monospecialized types, or types in which specialized species (the reindeer, yak and, finally, the llama, if not only pastoral nomadism is to be taken into account) were predominant, remained basically endemic.

118

3 The social preconditions of the relations between nomads and the outside world

Native model, theoretical model and reality The problem of how far the native (folk) model of social organization coincides with reality, what the functions of this model are, and so on, is not new in anthropology or, indeed, in other disciplines. Its practical significance is evident. The fact that stereotypes of social consciousness, of those notions, that is, which society has about itself and the day-to-day practical activity in it (la societypensee and la socie'te' vecu, in the language of Levi-Strauss) do not coincide is sufficiently well known. At the same time it is important for any scholar working on the social and political organization of nomads to establish their real character and structure. In the opinion of Spooner (1973:25-6) '. . . the basic function of the native model of social organization of a population is to allow its members to predict their relationship with each other from day to day by means of stable social groupings and stable identities, statuses and roles.' Such an opinion has specific foundations. However, one thing should be kept in mind. We should not think of the native model amongst nomads as a rigid structural model, joined to a reality which itself also changes little. The framework itself is changeable, and capable of accommodating both the static functioning of social forms and its transformations. It is no coincidence that nomads are patently aware of the structural principles of their models, but at the same time their sense of the boundaries and the individual details of them is frequently relatively vague. Finally, one other factor must be taken into account here. The native model is a component part of the ideology of a given society and therefore not only reflects the economic, social and political relations within it, but also influences them. It covers socialization, legitimization, normative factors, selection of personnel and a number of other aspects. For all that the reflection cannot be completely adequate (society cannot fully explain 1

1

Sec, for example, the discussion over many years of the tripartite theory of Indo-European widely in c o n n e d ion with llie works of I )iiiiu5zil. II')

Preconditions

of the relations between nomads and the outside

world

itself), and its influence cannot be completely spontaneous. In the ideology of any non-egalitarian society there are aspects which reflect the individual interests of specific groups, strata and classes; suffice it to mention here, if only by way of example, legitimization of power, about which anthropo­ logists interested in the origins of classes and the state are at present writing a great deal. The main peculiarities of the native model of many (although certainly not all) nomads are notions about society as an expanding family or minimum lineage, the descent principle and genealogy. More precisely, these are certain generalized peculiarities of many native models, a model for those models, their invariant. In effect, every society has its own model, and although there do exist models of one type, two absolutely identical ones are hardly likely to exist. Moreover, in nonegalitarian societies it is possible for several native models or submodels to exist at-the same time. Still in need of explanation are, on the one hand, how concretely the aforementioned characteristics of the native model correspond to diverse practice and, on the other hand, whether it is possible to create a single theoretical model of the sociopolitical organization of nomads. Anticipating the contents of the present chapter I am inclined immediately to reply to the second question in the negative. One can, of course, follow the many observers and scholars who think the sociopolitical structure of the most developed nomadic societies in the Eurasian steppes and in the Near and Middle East approximates to the following scheme: family - lineage or lineages of varying genealogical depth - clan - section or sections - tribe - tribal confederacy. However, such a model has more in common with a pile of bricks than a finished building and it scarcely brings us any closer to an understanding of nomadic society. First of all it must be pointed out that in this case we are dealing not with the native, but with the theoretical model. Nomads themselves are well aware that their social structure is segmentary, breaking up into more or less inclusive units at a number of different taxonomic levels; nevertheless the terminology they use to describe the corresponding segments is very un­ differentiated. This fact is partly due to the incomplete functional differen­ tiation and variability of the segments. Amongst the Arabs the term ashira (ashireh) is used to designate the most diverse segments of the social structure, from lineage where it corresponds to the term hamula (hamouleh) and many others to the higher-level segments. However, amongst different tribes in Northwest Arabia the sum total of several hamula has been called an ashira, but amongst other tribes it has been the other way round so that several ashira have made up one hamula (Jaussen, 1908:112f.; Bacon, 1958:120-30; Pershits, 1961:69-71). 120

Native model,

theoretical model and reality

In the Middle East this situation can be traced back to the Middle Ages. In Safavid Iran the same terminology from the most diverse origins - Turkic /'/ (el), Arabic ta'ifa (tayfa, tayfeh), Mongolian aimak (aimaq) was used to describe the same tribes and all their different subdivisions (Petrushevsky, 1949:94-5). The same inconsistency and interchangeability in use of terminology may be observed amongst many other nomads and seminomads (see, for example, Vladimirtsov, 1934:59; Bacon, 1958:58 on the Mongols in the Middle Ages; Zhdanko, 1950:78-9 on the Karakalpaks and Uzbeks; Abramzon, 1951:135, 139; Bacon, 1958:116 on the Kazakhs and Kirghiz; Markov, 1976:223 on the Turkmen; Trubetskoi, 1966:140-1 on the Bakhtiari; Barth, 1953:36, 37, 46; Avdal, 1959:152 on the Kurds; Pehrson, 1966:18 on the Marri Baluch; Asad, 1970:104 on the Kababish). Moreover, a nomad talking about his obok, ashira, tireh, hamula or qabila (qabileh), and a scholar translating these terms into theoretical language as clan, section or tribe are not only talking different languages, they are also thinking in different categories. For the nomad obok or ashira may signify one of the many descent groups to which he belongs and, depending on the context, they can be attached to units of a different taxonomic level. For the anthropologist clan or tribe are theoretical terms signifying a specific form of social organization which differs from other forms; they are terms which are inevitably linked with the burden of a specific tradition (and inertia) of thought. Since there is no generally accepted definition of these terms, nor any all-round understanding of the forms of social organization they define (evidently there never could be), the anthropologist must explain exactly what is meant by section, clan and tribe in the society which he is studying. If he then wants to go on and generalize he must examine the general functions of the clan and tribe amongst nomads. What then remains of the theoretical model itself, at least before this procedure is completed? One thing only. The social organization of nomads is complex and multi-levelled and is conceptualized in notions of kinship and descent. This fact in itself and everything else still need to be deciphered and explained. But even after this has been done many things are still unaccounted for: productive communities, territorial units, kindred groups, contract associations, etc. Even when an attempt is made to work out a structural-functional model of the social organization of a specific nomadic society, at the very least two or 2

3

2

Admittedly, it is not impossible that the non-differentiation and, particularly, the inconsistency of social terminology amongst nomads will seem exaggerated when we turn from relatively large ethnic groups as a whole to their separate subdivisions. A precise and single system of nomenclature is evidently more typical of the latter. Nevertheless, in this case the fact that social terminology either does not coincide at all, or does not fully coincide is remarkable. * In this case what interests me is not the dilemma of the emic/etic approach, but its results, for neither one or the other can be an end in itself.

121

Preconditions

of the relations between nomads and the outside

world

three different typological and partly taxonomic aspects or planes with different functions must be singled out beforehand. The first is the economic or productive plane; for example, the individual household - nuclear community - community of the second order - (community of the third order) - nomadic unit. The second is the genealogical plane; for example, the family, more precisely its unilinear core - microlineage - lineage macrolineage - clan - section - tribe. The third is the social, more precisely the sociopolitical which includes the territorial plane; for example, the family - family based primary kin group - sub-clan - clan - tribe - tribal association or confederation. Despite the fact that they are linked and interconnected, that separate segments coincide and that functions carried out by segments of different planes partially coincide, these planes cover distinct institutions and activi­ ties and their confusion, especially in descriptive works on social organiza­ tion, is inadmissible. Only after work on differentiating them has been completed (and with regard to nomadic societies this work has only just begun) and a specific model worked out, will it be time for an attempt to be made to single out certain invariants in similar models of the social organization of different nomadic societies. However, in this respect my aims are much narrower, for I am concerned with only some of the characteristics of the sociopolitical organization of nomads and only in so far as they are related to the main theme of this book. Unfortunately, therefore, I am sometimes unable to avoid confusing different typological aspects (something which I myself have only just condemned) and have to turn to individual blocks and bricks even, rather than to a completed building. On the other hand, I am proving by my own example that if criticism is a part of theoretical work then it is, undoubtedly, the easy part. Finally, one more very important characteristic of nomads must be mentioned here. The non-autarky, in many cases I would even say the anti-autarky of their economy, means that their social and political organization cannot be fully autonomous and that culturally to a certain degree they are not self-sufficient. The latter is determined both by internal and external factors, particularly by the nature of relations with the outside world, the degree of advancement and the specificity of the latter. It is no coincidence that nomads sometimes have more in common, socially and culturally, with agriculturalists in the same area than with nomads from other areas (Irons, 1979:362). None of this means, of course, that in the social organization of nomads there are no forms directjy linked to the economic specificity of nomadism, stimulated by the productive needs of the latter, and indirectly even to 4

4

Definitions of the terms used by the author are given further on in the text.

122

The problem

of ownership

in nomadic

societies

adaptation to a specific natural environment. Another factor must be taken into account. Both the sociopolitical organization of any nomadic society as a whole and many of its specific forms can and should be looked at from two angles: from within, as having risen directly out of the needs and particularities of the functioning of the society itself; and from outside, as having been stimulated, completely or partially, by the particularities and needs of its relations with the outside world. The problem of ownership in nomadic societies The economic relations which exist in nomadic societies are based on two important foundations: private ownership of livestock and corporative ownership of pastures. It can be said that with only rare and insignificant exceptions (see Pershits, 1959:40ff.) in all nomadic societies, both of the past and the present, the vast majority of livestock has belonged to separate individuals and/or their families. As a whole the wastage of manpower is less in extensive pastoralism than it is in agriculture; on the other hand, the task of looking after animals is more personal than is that of tending crops. Thus the productive specificity of pastoralism is such that right from the time of its emergence it has had to be based if not on private ownership, then on forms in which possession and use of livestock are particularly individualized (Khazanov, 1975:9, 93; Shnirelman, 1980). From the legal point of view amongst nomads private ownership of livestock is an indisputable right, but in practice in the majority of nomadic societies it is supplemented by different forms of reciprocity, and sometimes even by redistribution. The Bakhtiari of today are a rare exception (see Digard, 1973:48). The problem of the ownership of key natural resources in nomadic societies, first and foremost of pasture, is more complex. Soviet anthropo­ logists have been discussing this question for several decades. Certain scholars have suggested and continue to suggest that in the past, in the most developed nomadic societies, the ownership of pastures became, if not legally, at least practically, a question of private ownership or, more specifically, feudal in character (see, for example, Vladimirtsov, 1934; Potapov, 1954; Zlatkin, 1973; Lashuk, 1973; Fedorov-Davydov, 1973; see also Krader, 1979:230). Other scholars have disputed and continue to dispute this opinion (Tolybekov, 1959; Tolybekov, 1971; Shakhmatov, 1964; Markov, 1970; Markov, 1976; Khazanov, 1975; cf. Pershits, 1976). In my view we can only talk of a tendency for some kinds of pasture to become private property, and only with regard to individual nomadic societies. As a rule such societies are to be found in the modern period and in the present 123

Preconditions

of the relations between nomads and the outside

world

day, and to have integrated with sedentary societies into one system in which the nomads occupy a subservient position (on this situation in Africa in the present day see Dahl and Hjort, 1979:32-3). But even in these societies private ownership of pastures frequently involves a number of limitations and is much less complete and developed than it is in neighbouring sedentary societies. For example, in the Eurasian steppes the tendency for separate pastures to become the property of individuals emerged after nomads had been incorporated in the Russian and Chinese empires. However, it was never completely carried through. With the Kazakhs of the Inner (Bukei) Horde in the nineteenth century Khan Janger who, incidentally, had been granted the rank of MajorGeneral in the Russian army and had been raised to the noble estate, himself seized and distributed to his retinue pieces of land as private property (Tolybekov, 1971:388). Nevertheless, this case was exceptional. The law of 1868 broke definitively with the Russian government's former policy of backing khans and sultans, and all the land of the Kazakhs was declared state property (Viatkin, 1941:318). In Mongolia, in the nineteenth century and the beginning of the twentieth, land was looked upon as the property of the Manchurian government, property which Mongolian military-administrative units were granted the right to use. Maisky (1921:223,269) observed that'. . . the land is public property which, in fact, is presided over by khoshuns. Neither the princes, nor the highest lamas, nor Bogdo-Gogen himself have distinct pieces of land belonging only to them. Everybody pastures his livestock on an equal footing . . . With regard to where he can pasture the prince, as a general rule, has no privileges; his stock pastures with the stock of his subjects.' However, the control of the use of pasture set up by the Mongolian aristocracy considerably altered traditional norms. Back in the nineteenth century Dubrovo (1884:7) wrote of North Mongolia: 'The princes and taiji, i.e. the nobility, use the common land along with everybody else, at the same time they permit themselves to mark out the best pieces of this land and seize them for their own use, secure them and vigilantly watch to make sure nobody else uses the land which is now as it were their own. Infringement on rights to this land, seized without any agreement of the community, is severely punished.' A similar situation may be observed in the present day in parts of the Middle East (see, for example, Barth, 1964b on the Baluch). Nevertheless, corporative ownership of pastures is characteristic of the overwhelming majority of nomadic societies and, moreover, the forms of corporative ownership are far more varied than are those of the ownership of livestock by individuals and individual families. 124

The problem

of ownership

in nomadic

societies

At one extreme are relatively simple forms of corporative ownership, such as when land and pastures (the entire territory, that is) in theory are looked upon as belonging to the nomadic unit as a whole, and when in practice there is no fixed division (or only a vague division) of this territory between its different subdivisions. The best examples are provided by the nomads of East Africa (Gulliver, 1955:31; Spencer, 1965:5; R. and N. Dyson-Hudson, 1969:78, 79, 88) or by some of the nomads in the Sudan (Cunnison, 1966:27, 74; Asad, 1970:13). A similar situation may be observed amongst the Somali (Lewis, 1955:331-2), the Al Murrah (Cole, 1975:28, 95), the Marri Baluch (Pehrson, 1966:8) and others. At the other extreme are nomadic societies in which pastures are the supreme property of the entire nomadic unit, but in which, in practice, rights of possession and use of pastures are divided up between the different subdivisions of the society. This is, of course, only an outline, but it is founded on factual data; innumerable examples have been recorded amongst nomads in the Eurasian steppes and in the Near and Middle East. When writing about these regions scholars frequently describe specific instances of the dividing up of pastures, also wells, boundaries of routes of pastoral migration, etc. Moreover, ways in which rights of ownership, possession and use are divided up are often very complex. This form of corporative ownership of pastures is, to a considerable extent, linked to the regulating and control of their division and use. The carrying out of these functions involves the attribution of specific privileges to the ruling strata of nomads; such privileges are both social and material and particularly concern the pasturing of livestock in the best pastures. It is these privileges that some Soviet scholars cite as proof of the existence of the private ownership of pastures amongst nomads. However, despite isolated abuses, what we are dealing with in this instance is the carrying out of socially expedient managerial functions and their rewards, rather than rights towards private property. For example, the last Turkic qaghan had pastures which other nomads were forbidden to use. However, the herds pasturing on them had to serve as food during military campaigns. In the Middle Ages the quruks, that is those pastures which the Turks and Mongols were not allowed to utilize, acted as a kind of general reserve in the event of war and natural disasters (Barthold, 1963a:29, 34). In some nomadic societies ownership of wells and other watersources acquired an important independent significance. Again there were different forms of this ownership; some were more individual than others, which were corporative in character, or which were similar to corporative forms. As a rule, the more individualized the building and the maintaining of wells and the: more work this requires, the more individualized is the 125

Preconditions

of the relations between nomads and the outside

world

ownership of the wells in question. Sometimes it does not include the entire nomadic society, but only separate (usually the lower) subdivisions of the society. However, the individual or group ownership of wells can hardly be called the ownership of private property, because in no way does it signify that any members of the nomadic society have full ownership. In practice ownership is frequently limited to preferential use of the well and the right to immediate access, but not at all times and in all circumstances; it also allows other individuals and groups access to the well, usually without payment. In these circumstances a great deal depends on shortage of water. Thus rules can change from year to year and season to season (see, for example, Ishchenko et al, 1928:65 on the Kazakhs; Markov, 1976:224 on the Turkmen; Montagne, 1932:72; Lippens, 1956:184; Cole, 1975:32,86 on the nomads of Arabia and Syria; Stewart, 1973a:382 on the Moors; EvansPritchard, 1949:36 on the Bedouin of Cyrenaica; Gulliver, 1955:37; Spencer, 1966:6 on the nomads of East Africa). The different forms of corporate ownership of key resources in nomadic societies are explained by the joint operation of many factors. Some of them are principally linked to environment and climate (more or less regular precipitation and correspondingly stable pastoral migrations, uniformity or lack of uniformity of grazing territory, seasonal variations); others to economic factors (species-composition of herds and their pasturerequirements, nature of utilization of ecological zones and niches); and others, directly or indirectly, to the size of the given nomadic society and its social organization. Finally, even here the relations amongst nomads themselves and with the outside world should not be ignored. Family, household and community in nomadic societies In all nomadic societies there are two universal institutions - the family and the community. The presence of the family does not need to be proved since, obviously, it has always existed in all human societies. The question of the community requires special examination. First of all, however, something must be said about the family. Comparative materials show that, as a rule, nomadic families are not large and do not usually include more than two generations of adults. In the majority of nomadic societies nuclear families, consisting of a husband, wife and their unmarried sons and daughters, predominate. One variant of this kind of family which is widespread amongst the nomads of the Eurasian steppes, for example, is the patrilocal stem-family in which one of the married sons, usually the youngest, lives with his parents and inherits that part of their property which remains after the rest has been distributed among his elder brothers (for a detailed description see Khazanov, 1975:94-8). However, the stem-family is not only to be found amongst the 126

Family, household

and community

in nomadic

societies

nomads of the Eurasian steppes (see, for example, Stenning, 1959:4, 47 on the Fulani; Barth, 1964:12 on the Basseri). Except possibly in East Africa, in nomadic societies polygynous families comprise a relatively small percentage of the overall number of families. After the heads of these families die their families disintegrate, either into families consisting of the mothers and their children, or they partially join the ranks of other families in cases where the laws of levirate, according to which the son, brother or relative of a dead man inherits his wife and the obligation to take care of the children who are under age, operate. In the latter case the disintegration of the original family is only temporarily postponed until the children grow up and separate, making their own families. Sometimes there are specific reasons (a family with a large number of livestock needs to maintain its workforce or reluctance to break up the family herd, for example) why married sons separate only after their father's death. Sometimes brothers (usually the sons of one mother) continue to maintain the same household and remain joint owners of livestock, even after the death of their father. Other forms of extended families are also known. On the whole, however, all of them are only temporarily extended; they are not truly complex or extended families like the mir of the Eastern Slavs or the zadruga of the Southern Slavs, which had not been divided up for several generations at least and which consisted of scores and even hundreds of people (Kosven, 1963). Amongst nomads the break up of the family occurs sooner or later, usually in every other generation. In this respect all known forms of nomadic families are closer to nuclear than to extended families. Admittedly, assertions that extended families predominate in one or other nomadic society are not uncommon. However, an analysis of corresponding materials usually demonstrates that such assertions are founded on the confusing of two different institutions - families and primary kin groups. Moreover, the fact there is no generally accepted notion of what the extended family exactly is further confuses the issue. One characteristic of the nomadic family is that more often than not it coincides with a separate household. Consequently, amongst nomads the main economic feature of the family as a household is joint production (not excluding, of course, age and sex division) and joint consumption, meaning that each member has an undisputed right to a portion of the products produced in his household, and the presence of movable common property which is often at the disposal of or under the control of the head of the family. The family is a single and autonomous, ideally self-sufficient economic unit. Even in cases where a household and family do not fully coincide (as, for example, amongst pastoralists in East Africa), even where u husband, wife and children individually own livestock, whilst they make 127

Preconditions

of the relations between nomads and the outside

world

up one family general family interests and requirements, often embodied in the head of the family, have the upper hand, even though family interests may not always coincide with individual interests. Thus, amongst the Turkana the household is divided into as many yards as there are wives in the household. Each wife is allotted a certain number of stock and she separately prepares food for herself and her children. But when necessity demands the wives share their reserves of food, meat and blood is divided up amongst all the members of the family and the head of the family has the final say in the apportioning of livestock (Gulliver, 1955:128). Strange as this may seem, the primary kin group (patronymy, large extended family, microlineage and even clan, according to the terminology of different scholars) is one form of social organization on which very little comparative work has been done, although in practice scholars researching into primitive and traditional societies are constantly com­ ing up against it. What I understand by primary kin group is a number of separate and independent families which are very closely connected with one another through ties of kinship, reciprocal relations, common residence, etc., the core of which is made up of very close consanguines (microlineage), who are descended from one close ancestor and who in the past have frequently made up one family - brothers, cousins, uncles, nephews, etc. Amongst nomads a primary kin group consists of closely related families which all year round, or for part of the year, pasture together and help and support each other. Such groups are, of course, considerably less stable than an individual family and all the families of a primary kin group run their own households and keep their own livestock. The primary kin group consists of several autonomous economic cells which do not automatically have the right to make claims on each other's property and labour. Thus the primary kin group must not be confused with the extended family. Nicolaisen (1963:146) writes that amongst the Tuareg of Ahaggar there exists the '. . . extended family which co-operates in goat-herding and in various other tasks, although all animals are individually owned and every tent or household is self-supporting - "It has its own churning bag and its own millet bag", as the Tuareg put it.' Essentially what he is describing here is a primary kin group. Gulliver (1955:156) writes of 'extended families' amongst the Turkana, but in fact what he is describing are primary kin groups; he also observes that the basic economic cell in Turkana society is the nuclear family, consisting of a husband, his wives and children, in which only some elements of the 5

5

In a functional-structural analysis it is extremely important to differentiate between exogamic and non-exogamic primary kin groups. However, if I were to do that now it would lead me away from my main theme.

128

Family,

household

and community

in nomadic

societies

polygynous family and the patrilineal extended family are to be found because sons only become fully independent after the death of their father (cf. Gulliver's description of the correlation between family and primary kin group, 'the extended family', amongst the Jie - 1955:76, 81). In his description of the extended family amongst the Humr Baggara, Cunnison (1966:60) writes that this usually consists of the sons of a dead father and their wives and children. But what he describes is a primary kin group in the sense in which I have described the term rather than a family, for amongst the Humr Baggara man's '. . . primary responsibility is for his own nuclear family; responsibility for his extended family runs a close second' (cf. Stenning, 1958:93; Frantz, 1978:105, 106 on the Fulani). At the same time the differences between the family and primary kin group are clearly pointed out by Barth (1962:344-5; cf. 1964a:74) amongst the nomads of Southwest Asia, by Pershits (1961:69) amongst the nomads of Arabia, and by Bessac (1965:383) amongst the Mongols (see also Swidler, 1973:27; Glatzer, 1977:130). From the time of their emergence in the Eurasian steppes and throughout their entire existence as nomads, the nuclear family and a variant of it, the stem-family, were predominant amongst the nomads of the Eurasian steppes (Khazanov, 1975:73-6). According to Rashid al-Din (1960:107) it was the ancient custom amongst Mongols and Turks for eldest sons to be singled out in the lifetime of their fathers and given goods, livestock and a herd of sheep. Alongside such families there were always primary kin groups. Members of the agnatic core of the latter frequently called themselves 'sons of one father' (cf. the Kazakh ata balasy, Kirghiz bir atanyn baldary and Turkmenian bir ata) although in reality more than just brothers were included in this core. In this instance the word 'father' was used in the sense of 'a very close ancestor' (Abramzon, 1951:153). It is these groups which are regarded by Bacon (1958:57, 68-9, 84), Krader (1963:370-1) and Szynkiewicz (1975:115-16) as extended families. At the time when I was studying the social organization of the ancient nomads of the Eurasian steppes, I noticed that according to archaeological data it would seem that as these nomads became more and more sedentary so the size of their families and the number of generations in the families increased. Consultation of comparative ethnographic materials showed that a similar phenomenon can be traced in the Eurasian steppe region right up to the modern period (Khazanov, 1975:73ff.). It has also been recorded in the Near East (Jaussen, 1908:llff.; Pershits, 1961:89) and the Middle East (Swidler, 1973:37). A curious series of patterns emerges. Within one cultural area the size of the nomadic family is usually smaller than the semi-nomadic family, when these families are of comparable social and economic standing. There are, evidently, economic explanations for this. The workforce required for 129

Preconditions

of the relations between nomads and the outside

world

production in pastoral nomadism on the whole is less than that which is required by agriculture. Shortage of labour in a household of average prosperity is usually seasonal or temporary and linked to a specific domestic cycle. In part it is compensated by the institutions of reciprocity among nomads. It is possible that these differences have been further influenced by the specificity of livestock as a form of property which does not require capital investment and prolonged and intensive outlay of labour, which is movable and can be divided without difficulty, which is liable to sharp fluctuations in numbers and at the same time requires extremely individualized care. In traditional agriculture, particularly when based on irrigation, the need for a workforce is far greater. The desire to slow down the process of the division of immovable property, land, could also have played a specific role. I think it worth pointing out, in conclusion, that when hunters and gatherers become sedentary the size of their families also increases (R. Cohen, 1978:42-3). Now we can move on to the community. Both in Soviet theoretical literature and in the theoretical literature published in English this term ranks among the most polysemantic. (For a list of various definitions see, for example, Webster, 1960, s.v.; Murdock, 1965:79; Khazanov, 1975b:3.) Nevertheless, in many, if not in the majority of peasant societies a community constitutes something more than a set of people living in face-to-face relationships and isolated by a kind of gap in social relationships from other similar groups. In such societies, alongside its social functions, the community has clearly defined economic and even productive functions which are linked to the interests and requirements of the immediate producers. There are many variants of the peasant community to be found in different regions, different periods and in connection with different economic structures. Nomadic communities share a number of common features with them, while at the same time each have their own individual characteristics. Neither from the point of view of property or production are any separate families and households in nomadic societies identical. This is because each one has a different number of livestock and a different number of available working hands, both of which are variable. The problem is further complicated by the fact that nomadic families are usually relatively small and that the workforce required by the household is subject to seasonal variations, quite apart from variations linked to the domestic cycle. As a rule it is impossible, or at least very difficult, for one nomadic household of average prosperity to accomplish a complete productive cycle (let alone several cycles in succession) on its own. Although nomadic households are not as self-sufficient as peasant 130

Family, household

and community

in nomadic

societies

households, they are like the majority of peasant households (particularly of pre-capitalist times), and unlike capitalist farming households in that they cannot maintain production without labour co-operation and other forms of mutual aid. Although there are many such forms of aid, production amongst nomads and peasants is impossible without some kind of established system in those relations, which constitute the foundation and essence of the community, having been worked out. In comparison to those, other factors contributing to the emergence of a nomadic community (attitude to key resources, for example) are secondary. Wherever small peasant households are to be found, a peasant community emerges. Wherever small nomadic households are to be found, a nomadic community emerges. The essential characteristics of communities in different nomadic societies are, of course, variable, and the significance of communities within general systems of social relations is variable. The functioning of nomadic communities is to be observed more clearly than anywhere else amongst the nomads of the Eurasian steppes, and the Near and Middle East (Pershits and Khazanov, 1979). In the majority of societies belonging to the types of nomadism which have been examined, one main feature of the community is its relation to key resources. Usually the community only uses these resources according to the extent of its members' participation in a nomadic unit or a subdivision of one, in all of which proprietary rights are contained. It is not only the ownership of key resources, particularly of pastures in many nomadic societies, which is shared; rights to use them are frequently incomplete and have to be shared. Thus in the Eurasian steppes pastures, especially summer pastures, may be used by several nomadic groups from one larger unit. Sometimes these groups would appear to be independent communities and sometimes component parts of a community of a higher taxonomic level. For example, of the Adai Kazakh it has been recorded '. . . in their summer camps the auls for the most part pasture together, particularly within the main clan groups' (Ishchenko et ai, 1928:108). Amongst the Kirghiz winter camps were allotted more or less to communities consisting of small groups of close relatives (ails). But summer ails of one subdivision pastured their stock in communally owned grazing territory. However, sometimes the routes of the migrations of several subdivisions crossed and in this case the adjacent land was looked upon as common territory (Ploskikh, 1972:27-9). Such divided ownership of property also may be observed fairly frequently in sedentary societies, for example in antiquity (Graeco-Roman) or in feudal society. In these societies rights of ownership, possession and use were at least fairly clearly demarcated on the horizontal level; but in 13J

Preconditions

of the relations between nomads and the outside

world

nomadic societies such rights remain extremely vague and are often the source of internal conflicts. Another main feature of the nomadic community which is inextricably linked with the one already mentioned consists in its distinctive vertical structure. From the point of view of production a single community comprises a small group (rarely of more than 15-20 people in the Eurasian steppes and the Near and Middle East) of independent households which move together all year round or for part of the year, which may jointly pasture livestock, and/or which to a very great extent are linked by other ties of mutual aid, and sometimes even of mutual responsibilities and for mutual defence. One early observer (Levshin, 1832, pt. 111:24) wrote: 'The Kirghiz [i.e. Kazakhs - A.K.] rarely roam in great numbers in one place, for then their herds are crowded; but associations consist of several families which are connected by kinship or mutual need, and they move together from one camp to another and do not separate without specific reason. This mobile village they call an aul; the number of kibitkas in an aul depends on individual circumstances.' Kazakh auls consisted of 2-4 households, rarely of more than 8 (Semeniuk, 1973:44) right up to the enforced collectivization and sedentarization of the thirties of this century (Briskin, 1929:11). In the thirties communities consisting of 5-6 households were most widespread amongst the Mongols (Simukov, 1933:23). Sometimes such communities of the primary order, which may be called nuclear nomadic communities, coincide with primary kin groups. But this is not always the case. Frequently they consist not only of agnates and their families, but also of affines, cognates and individuals connected only by distant kinship, and sometimes not even that (see, for example, Zimanov, 1958:77 on the Kazakhs; Cole, 1975:48, 63 on the Al Murrah; Barth, 1964:21 on the Basseri; Pehrson, 1966:72ff.; Pastner, 1971:75 on the Baluch; Lewis, 1965:333-4 on the Somali). It is noteworthy that nomads themselves sometimes single out nuclear communities, and sometimes give them special names: aal and khot amongst the Tuvinians, ail or aul amongst the Kirghiz and Kazakhs, khoton amongst the Kalmucks and dar amongst the Al Murrah. Considerably more often, however, because social terminology is usually polysemantic nomads describe such communities by terms which also are used for descent or kin groups, for example hamula amongst the Arabs or oba amongst the Turkmen (cf. Tapper, 1979:59). In many nomadic societies communities of the primary order join together in a community of the second order. The members of the latter community use the same pastures and/or water-sources at specific times of year, join up for pastoral migrations, and are linked by various social, kin and other ties, sometimes also by specific forms of mutual aid. Finally, in 132

Family, household

and community

in nomadic

societies

some cases several communities of the second order can make up a community of a third order with corresponding rights to natural resources, supplemented by various other ties. Thus, amongst the Kalmucks, the lowest level of the community was the khoton which was made up of closely related families roaming together. Kin khotons united to form the terl (terlmud) and arvan, the members of which strove to remain in fairly close proximity to one another. The third level consisted of the anghi and aimak, the lands of which specifically were called an otok, a district, that is, with more or less clearly defined borders (Nebolsin, 1852:7-8; Erdniev, 1970:122, 209-11). Three such levels of community organization are described by Pastner (1971:175-6, 180) amongst the Baluch in the Makran region of Pakistan. The terms Pastner uses for these levels are: the nomadic community, consisting of 2-4 nuclear families, the micro-pastoral orbit, consisting of up to 10 camps (approximately 200 people), and the macro-pastoral orbit. What Evans-Pritchard (1949:56; cf. Peters, 1967:262; Peters, 1960:31) calls extended family groups (biyut) amongst the Bedouin of Cyrenaica are, in fact, different levels of a nomadic community. This becomes clear in his description of them: 'Their members live in the same stretch of tribal territory, move during the rains to the same grazing grounds, use the same wells during the dry season and cultivate adjacent strips of arable land. The members of a biyut have a lively sense of solidarity . . . They are jointly responsible for a wrong any one of them may commit. The smaller biyut are often identical with camps, usually from five to ten tents, in grazing grounds, and several closely related biyut camp near together in the vicinity of springs and wells in the summer, their combined camps then amounting sometimes to over a hundred tents.' Amongst the Tuareg of Ahaggar a nuclear community consisted of 2-7 families. In a favourable season several such communities joined to form a secondary community of 10-20 families. This was a co-operative unit in which the pasturing of camels and caravan trade, for which the workforce of an individual nuclear community was too small, were joint enterprises (Nicolaisen, 1963:146). The third main feature of the nomadic community is fluctuation of members which, on the whole, is considerably greater than that of peasant communities, or at least of the majority of the latter (cf. Spooner, 1973:14-15). In contrast to many peasant communities, the members of nomadic communities do not live, or in this case do not move together permanently. For different reasons, season, size of herds, productivity of pastures and the state of water-sources, members of a nomadic community can move together or separately. In effect, it is by no means the case that the different families of nuclear nomadic communities always pasture together. The impossibility of herding large numbers of livestock together is 133

Preconditions

of the relations between nomads and the outside

world

mentioned in the Bible where it is told how Abraham and Lot separated their herds (Genesis 13.6-11). Apart from this, the balance which was described in Chapter 1 does not only relate generally to nomadic society as a whole, but it also affects individual households, for example, in different stages of their domestic cycle. This second aspect of the balance is linked to short-term microvariations and involves corresponding variations and changes in the size, composition and structure of nomadic communities. Economic problems, the need to maintain and strengthen both kin and contractual social ties and other factors have also influenced the size and fluidity of nomadic communities. The political aspect of their functioning must also be taken into account. In particular, the fact that nomadic units are weak and decentralized and the absence of reliable guarantors of safety in the form of the state or a proper military-political organization can sometimes encourage the ten­ dency of communities to integrate. On the other hand, relative security can lead to their disintegration. In the twelfth century, which for the Mongols was a period of constant internecine feuds, the Mongols frequently moved in xuriens (Rashid al-Din, 1952a:86). The discussion about what exactly these xuriens were has already been going on for several decades. It is my view that in one of their functions they sometimes were communities of the second order, but within these communities there was social differentiation and their composition was very complex. As internal stability was established amongst the Mongols, the ail way of moving, that is in nuclear communities, began to prevail. In the eighteenth century a considerable number of Kazakh auls chose to move together on account of the unsettled political conditions surrounding them (Tolybekov, 1959:206). But such associations of several nuclear communities were temporary and collapsed when political conditions were calm (Zimanov, 1958:70-1). Of course, the divided ownership of key resources and the vertical structure of communities themselves facilitated movement between them and the formation of new communities. It is no coincidence that the majority of such transitions took place within communities in the higher taxonomic order. For example, amongst the Kazakhs in the nineteenth century it was not uncommon for members of one and the same subdivision who jointly used (or possessed) a specific pastoral territory to form several nomadic communities, transferring relatively easily from one to another (Shakhmatov, 1964:66ff.). The same is recorded by Sweet (1965:137) of the nomads of Northern Arabia, amongst whom groups founded on minimal lineages (i.e. nuclear communities), individual families and even brothers may break away from their kinsmen and move separately or together with other groups. 134

Family, household

and community

in nomadic

societies

Such a situation is linked with yet another feature of nomadic communities which, incidentally, is also found in certain non-nomadic societies (on this point see Khazanov, 1975b:6; Pershits, 1978:155). This is the fact that from the structural point of view communities are frequently based on ties of kinship and common descent. At any rate individuals linked by these ties usually make up the core of nomadic communities (cf. Tapper, 1979:59). If a nomadic community is founded on ties of kinship and descent, does this not then constitute a distinct institution, one of the social or genealogical subdivisions of the nomadic society? The answer is essentially no. Despite its overlay of kinship and common descent, the essence of a nomadic community consists first and foremost in neighbourliness and production. Thus private ownership of livestock and independent production by individual households form the economic bases of the nomadic community. It is curious that sometimes even co-operation and mutual aid in the nomadic community seem to follow two courses. One is the neighbourly course and is concerned primarily with the needs of production; the other is based on kinship (for example, payment of bridewealth, participation in blood feuds, etc.) and is more connected with social functions. Thus it is not always the same group of individuals which has to be involved in both forms of mutual aid. Individuals from different communities, for example, may participate in mutual aid based on kinship. This may be very clearly observed amongst the nomads of Northern Arabia. Here it is thought that all neighbours are obliged to help out with such work as sheep shearing (Musil, 1908:285). However, assistance in the event of festivals of circumcision, marriages and, particularly, payment of blood dues is thought to be in the main the business of kinsmen, particularly of khamsa, i.e. members of one of the kin groups (Hess, 1932:32; Musil, 1928:48; Daghestani, 1932:202). Anthropologists have singled out and described nuclear nomadic communities, to which they have given various names, amongst many nomadic peoples. For example, Evans-Pritchard (1949:56) and Peters (1960:35) write of 'camps' (tertiary segments) amongst the Bedouin of Cyrenaica; Nicolaisen (1963:146f.) of 'camps' amongst the Tuareg; Torry (1976:272) of the 'homestead' amongst the Gabra; Awad (1962:333) of 'bands' amongst the nomads of the Middle East; Barth (1962:345) of 'co-operative herding units' amongst the nomads of Southwest Asia; Barth (1964:25) of 'herding units' amongst the Basseri; Pehrson (1966:71) of 'camps' amongst the Marri Baluch; Salzman (1971:186) of 'local communi­ ties' amongst the Shah Nawazi Baluch; Swidler (1973:27, 30) of 'camps' amongst the Brahui and the Baluch (cf. 'the subsistence grouping', Spooner, 1973:26 or 'the local exploitation group', N. Dyson-Hudson, 1972:11). However, it is often forgotten that usually such nuclear communities are 135

Preconditions

of the relations between nomads and the outside

world

only the autonomous, and most noticeable, component part of wider nomadic communal systems. An opposing view to that one is to be found in an interesting article by Tapper (1979). His 'herding units' essentially coincide with the nuclear community. Tapper (1979:49) suggests, and I am in agreement with him, that the 'herding unit' of nomads is first and foremost an economic institution. However, in his view a 'basic nomadic community' is primarily a comparatively stable 'primary sociological unit', 'basic political unit' and he only calls a community that which I look upon as a community of the second or third order. For example, he writes (1979:50) that amongst the Bedouin Al Murrah the community is '. . . a summer camp group which gathers around its own wells.' But Cole (1975:42, 63), on whose data Tapper's conclusion is based, also singles out the ddr, which is essentially a nuclear community, although it is an unstable one in which members move together for part of the year; the ddr consists of several households (bayt) which pasture their stock separately, but which eat together and are linked by ties of mutual defence which are stronger than consanguineal ties. In exactly the same way, the camps of the Basseri which Tapper (1979:54), in agreement with Barth (1964:25), considers to be 'the primary communities of nomadic Basseri society' are, in my view, communities of the second order. The primary (nuclear) community is the 'co-operative herding unit', an association of 2-5 households for the joint herding of livestock (Barth, 1964:21-3, 42). It could appear that the problem of what is a primary and what is a secondary community is far-fetched, like the discussion of Swift's Big-Endians and Small-Endians. But I do not think that this is the case. The primary community usually emerges first and foremost from the require­ ments of production, even if they are not always constant ones, expressed more or less explicitly and which manifest themselves in different ways. At the same time communities of the second and third order are functionally more varied. In some cases they (or their core) are founded on the joint possession or use of key resources, in others mainly on the fulfilment of social (and political) functions. Although the community exists in other types of nomadism, it does so less noticeably than in the types examined above. > For example, amongst the Kababish and the Humr Baggara of the Sudan only the primary community is to be found, but its composition is very unstable and can change every year (Asad, 1970:20,128; Cunnison, 1966:9, 42, 59f.). At the same time the nuclear community is clearly observable amongst the Fulani. Its core comprises an 'agnatic lineage group' consisting 6

6

This shows how co-operation amongst nomads can assume different forms,

136

Family,

household

and community

in nomadic

societies

of no more than three generations (about 20 families, which is about 100 people). On the whole a nuclear community coincides with a primary kin group, although it does not necessarily entirely coincide. The community has well established rights to pastures and water-sources, its members move together all year round or for most of it and are linked by ties of co-operation and mutual aid. Such communities join to form communities of the second order and move together in the rainy season (Stenning, 1959:38, 41, 50, 52-3; Stenning, 1965:367 , 378-9 , 383). Amongst the Gabra, who as a whole may be included in the Northeast African sub-type of nomadism of the Near Eastern type, the three levels of the community may be observed (Torry, 1976:272-80). Amongst the nomads of Tibet two levels of community stand out clearly and, moreover, the nuclear community often coincides with a primary kin group (Ekvall, 1968:28). The community is much weaker in the nomadism of the East African type. This is because ecological conditions in the area are such that every household must be able to use, at its own discretion, all the pastures in the territory belonging to the nomadic society of which it is a part. However, apart from a system of social ties for economic security, there is here mutual aid in production between individual households based on contractual ties of a kin and/or partner type, although these ties are not constant. Thus amongst the Samburu the community exists in the form of 'small settlements', consisting of 4-10 households of which the majority are linked by kinship. The composition of these communities may change after every pastoral migration, but all the households constituting the communities are linked by ties of productive co-operation and social aid, without which none of them can survive (Spencer, 1965:6, 8, 9, 12, 15, 16, 20, 22-3; Spencer, 1973:24—5). Amongst the Karimojong there are changeable stock camps, associations which emerge for 'security and sociability' (R. and N. Dyson-Hudson, 1970:110). Nor can it be denied that the community exists amongst the Turkana, although a household rarely spends an entire year in one community. Several levels of the community may even be discerned here: 'the primary neighbourhood' (2 or 3-5 households), a co-operative producing unit; 'the secondary neighbourhood' (5-20 households), 'based on a common water-point, often using common pastures to the temporary de facto exclusion of the other people'; and even 'the tertiary neighbour­ hood', based on certain social ties (Gulliver, 1955:11-12, 33, 37, 43). Amongst the reindeer-herders of Northern Europe the community may easily be observed (Brodnev, 1959:75; Vorren and Manker, 1962:143-4; Obshchestvennyi stroi u narodov Severnoi Sibiri, 1970:405-12; Vasilev, 1976:331-2; Krupnik, 1976:63-4). Amongst the Nentsy and Chukchi at the cad of the nineteenth century and the beginning of the twentieth there were 137

Preconditions

of the relations between nomads and the outside

world

two levels of the community; amongst the Lapps there was one, although there are indications that in spring and summer nuclear communities in certain groups had '. . . a tendency to coalesce into larger units', i.e. communities of the second order (Whitaker, 1955:62). The nuclear community (the nomad camp, what the Lapps call sii'dd and the Nentsy call parma) consists of an association of 2-5 households (10-25 people) for the joint pasturing of deer. Joint pasturing frequently occurs because of ecological conditions, as well as for straightforward producing needs. In the summer when the animals are plagued by blood-sucking insects, gnats and gadflies, a large herd is easier to protect than a small one because the animals in the former do not scatter, rather they cluster together and move around as such (Obshchestvennyi stroi u narodov Severnoi Sibiri, 1970:396). Amongst the Nentsy a group of neighbouring nomad camps is usually, although not always, linked by ties of kinship and affinity and mutual aid, and constitutes a community of the second order (100-250 people and more). Within these communities transitions from one nomad camp to another are not uncommon. Amongst the Chukchi the nuclear community, nomad camp, sometimes coincided with the varat (literally 'a gathering of people living together'), an association of 10-15 families. But in other instances the varat was a summer association of several nomad camps (Dolgikh and Levin, 1951:102-3). Kinship and descent in nomadic societies In Chapter 110 of Shih Chi, which is devoted to the Hsiung-nu, Ssu-Ma-Ch'ien tells an interesting story about Mao-tun, founder of the state of the Hsiung-nu. The Tonghu who were neighbours of the Hsiung-nu, asked Mao-tun to give them his best horse and then they asked him for his favourite wife. Contrary to the advice of his followers Mao-tun did as he had been asked for he wanted relations between the two peoples to remain peaceful. But when the Tonghu asked the Hsiung-nu for territory which lay fallow and Mao-tun's followers advised him to grant that wish, Mao-tun was angry and said: 'Land is the foundation of the state, so how can you give it away.' And he executed all those who had wanted him to relinquish land (quoted in Taskin, 1968:38-9). The story is probably a legend, but it is still a very instructive one. Nomads are well accustomed to the notion that territory belongs to one or other unit. The specificity of territorial links in nomadic societies only consists in the fact that they are not straightforward, rather they are settled and realized by ties of kinship (pseudo-kinship). It is, evidently, the mobility of nomads which limits the development of direct territorial and neighbourly links, thus 138

Kinship

and descent in nomadic

societies

leaving kinship as the best alternative for the expression of social relations. The mere statement of the fact that everywhere (or almost everywhere) a configuration of nomadic societies is founded on kinship (cf. Spooner, 1973:28) does not, of course, mean very much. Firstly, it is not only in nomadic societies that relations of kinship (pseudo-kinship) form the structural basis of social organization, they do likewise in many other societies in which there are different economic systems. Secondly, even a general configuration of societies founded on kinship is defined only to a certain extent by kinship, while considerable freedom is left for specific forms of sociopolitical organization. Strictly speaking, this is the sort of picture of nomads which emerges. Consequently, with regard to nomads a more precise definition is required. The mobility of nomads and the permanent instability of pastoral economy give rise to a fluid social organization, which is capable of change and which has the requisite segmentary means with which to accomplish this. In the majority of nomadic societies the most appropriate principle for this turns out to be descent, in which the structuring role of the relations of kinship is conceptualized. Thanks to the application of the principle of descent, the complex and multi-level character of the social organization of nomads is often expressed in the aggregate of separate segments, discrete descent groups of different genealogical depth which, if the need arises, are capable of fission or fusion. However, segmentation itself is not necessarily connected with notions that common descent embraces all the members of a given society. However, the assertion that the principle of descent is the most important factor in the structuring of the social and political organization of nomads needs considerable qualification. Firstly, for the most part this assertion is only applicable to pure nomads; amongst semi-nomads and, even more so, amongst extensive pastoralists the picture may sometimes be different. Secondly, the principle of descent does not exclude, nor can it exclude a direct realization of relations of kinship. I have already remarked on the fact that not all forms of social organization amongst nomads acquire the form of descent groups, and not all of them are constructed on the basis of a single principle of descent. Even leaving aside the controversy between native model and reality, between ideology and social practice, it may be said that in the societies in question kinship and descent not only operate primarily in different spheres and on different levels, but in part their functional directionality is different. Ibn Khaldun clearly recognized these differences when he wrote that asabia ('asabiyah) based on real kinship, was a considerably more effective unit­ ing force than general belonging to a tribe (Ibn Khaldoun, 1863, i:275). 'The reason for this is that the purpose of group feelings, which is defence and 139

Preconditions

of the relations between nomads and the outside

world

aggression, can be fulfilled only with the help of a common descent. For . . . blood relations and other close relatives help each other, while strangers and outsiders do not' (Ibn Khaldun, 1967:147-8). Units at lower levels of segmentation which, first and foremost, are connected with social, economic and more narrowly productive needs rely on kin and contractual relations (on the role of the latter see, for example, Spooner, 1973:25-6; Marx, 1977:356, 357; Marx, 1978:61f.). The higher subdivisions, the functions of which are primarily sociopolitical and in part include their functions as guarantors or holders of corporative proprietary rights to key resources, are based on the principle of descent (Spooner, 1973:24; Barth, 1973:18). But this is only one side of the picture. Kinship regulates relations within a relatively small collective (group) of people; it mediates the individual's position in a system of horizontal ties by superseding the discrete character of different descent groups. Descent regulates relations between different groups and at the same time establishes the individual's membership in a given society as a whole and in specific subdivisions of it; this membership involves both corresponding rights and commitments and sometimes even social positions. Kinship establishes the position of the individual in society, descent legitimizes it (cf. Marx, 1977:358-9). Descent operates in nomadic societies in two ways: in governing the real allocation of genealogies (genealogical descent), and in providing in the notion of common descent a bond for all the members of a given society. It need hardly be said that the two principles do not always coincide in their operation. Although the functions of the two are not entirely separable, they do not coincide in every way. Genealogies are important in the forming of descent groups and in defining how they are interconnected. The notion of common descent provides a theoretical foundation for social integration. Thus the Afghans all look upon Kais (Abdul Rashid) as their ancestor; his sons, grandsons and great grandsons are regarded as the founders of all the most powerful Afghan tribes (Elphistone, 1819, i:248,253; Reisner, 1954:52 n.2; Spain, 1963:41; Glatzer, 1977:107ff.). All the Somali trace their ancestry to Aquil Abuu Taalib, cousin of the Prophet and brother of Ali (Lewis, 1961:12), the Turkmen to Oghuz Khan (Abu-l-Gazi, 1958:40), the Buryat to Bargu-Baatur (Humphrey, 1979:247). See also Cunnison, 1966:111 on the Humr Baggara; Evans-Pritchard, 1949:55; Peters, 1960:29 7

7

Admittedly, Ibn Khaldun recognized just as clearly the relativity of such kin ties: 'Client relationship and contacts with slaves or allies have the same effect as [common descent]. The consequences of common descent, though natural, still are something imaginary. The real thing to bring about close contact is social intercourse, friendly association, long familiarity, and the companionship that results from growing up together, having the same wet nurse and sharing the other circumstances of death and life. If close contact is established in such H manner, the result will be affection and cooperation' (Ibn Khaldun, 1967:148).

140

Kinship and descent in nomadic

societies

on the Bedouin of Cyrenaica; Salzman, 1971:186 on the Shah Nawazi Baluch. However, the Samburu clearly acknowledge the different descent of separate clans and sub-clans (Spencer, 1965:77), and it is noticeable that links between these subdivisions are fairly weak. Amongst the Fulani the notion that all members of the clan are descended from one ancestor does not exist (Stenning, 1959:53-4). A good example of how ideological needs encourage the notion of common descent is provided by the Kazakhs. The Mongols destroyed the social organization of the Turkic speaking nomads who are regarded as the principal ancestors of the Kazakhs; they reshuffled the subdivisions amongst those nomads and caused considerable ethnic regrouping in the Eurasian steppes. But as soon as the nomadic groups of different descent which were to form the Kazakh people stood united politically and, following on from this regained self-identity, they felt the need to trace their descent back to a common ancestor who, admittedly, had still to be generally recognized (Levshin, 1932, n:25-31; Kharuzin, 1889:26-7; Grodekov, 1889:2-3; Aristov, 1894:394-7; Chuloshnikov, 1924:199; Abramzon, 1951:138; Bacon, 1958:67). However, for a number of reasons, primarily the weakness and instability of Kazakh polities, work on genealogies fell behind. As a result, a single genealogical system to establish common descent amongst all the Kazakhs was never created. In the Eurasian steppes in the post-Mongol period Turkic-speaking tribes would often consist of an original Mongolian speaking core, onto which separate subdivisions of Turkic speaking tribes adjoined. In time only the appellations of the tribes remained Mongolian. Such, for example, is the history of the well-known Konghrad (Kunghrad) tribe of Uzbeks. Nevertheless, all members of a new unit traced their descent to one common ancestor. A similar picture may be observed amongst certain nomadic tribes in the Near East (Musil, 1908, m:112f.; Philby, 1922, n:176; Pershits, 1961:86-7). Furthermore, the notion of common descent could be of significance in yet one more sense, which is here tentatively referred to as the external-political. It was able to sanctify relations between different nomadic unities by establishing the existence or absence of quasi-kinship between them. After the Mongol conquests the lives of the Turkic-speaking and Mongolian-speaking nomads became closely intertwined, and so the idea of their common descent emerged. It was recorded by Rashid al-Din (1952a:76, 77; cf. Bacon, 1958:55), but it scarcely was only an invention of this scholar-politician. According to the Rwala, all the Anaza tribes have I he same ancestor, thus they regard them as their paternal cousins, beni al-'umm (Jaussen, 1908:111-12). It seems to me that when Salzman (1978c:627) asserts t h a t ' . . . lineage 141

Preconditions

of the relations between nomads and the outside

world

ideology is, during times of stability, a social structure in reserve', he is somewhat limiting both its potential capabilities and its practical applicabil­ ity. The genealogical principle is also extremely useful to nomads for reasons other than the forming of descent groups; in essence it is polyfunctional, which explains why it is so widespread and vital. Genealogies can be consciously or unconsciously manipulated, they are capable of broadening and narrowing, and of splitting up and merging in accordance with practical necessities and a specific historical situation. In such cases nomadic social organization as a whole and its various levels acquire the ideological flexibility they need and the ability to re-organize to suit new conditions, without losing the structuring principles. One characteristic feature of the Turkic-speaking nomads of the Eurasian steppes was that the number of vertical segments in their genealogical line did not remain constant. The genealogical structure was constantly being reformed, which is a problem that scholars have frequently come up against, once they have established divergences in their genealogical diagrams of the very same tribes. Amongst the Humr Baggara the majority of subdivisions, from minimal lineage to the sections into which the tribe is divided, can easily change their position in the general structure, and the structure easily adapts itself to such changes (Cunnison, 1966:10). Many examples are to be found in the relevant literature of the significance which is attached to genealogies in many nomadic societies. Rashid al-Din (1952a:29) wrote: 'The custom of Mongols is such that they preserve the genealogy of [their] ancestors and teach and instruct with [knowledge of] genealogy each child who is born into this world . . . Not amongst one of the other tribes, excluding Mongols, does this custom exist, only amongst the Arabs who [also] preserve [within memory] their descent . . . Thus amongst them [i.e. Mongols - A.K.] there is no religious community, no faith with the help of which they would set a child, like other [people], on the righteous path, for to each of their children a father and mother explain the tradition of clan and genealogy.' Every Somali knows his genealogy, which numbers 20-30 generations (Lewis, 1961:128; Lewis, 1962:40). One other function of genealogies is that they legitimize social inequality in native models which are already heterogeneous. (This aspect is well researched by Humphrey, 1979, in her study of the Buryat.) Thus, although all the Scythians thought they were descendants of Targitaus the descent of the various uneven subdivisions of their society was linked with the various sons of the hero primal ancestor (Herodotus, IV.5-6; see also Khazanov, 1978:426). Frequently it is the aristocracy in a society which cultivates knowledge of genealogies and manipulates them so as to give an ideological basis to their ruling positions (see, for example, Stewart, 1973a:379 on the 142

Kinship and descent in nomadic

societies

Moors). Grodekov (1889:12) wrote of the Kazakhs and Kirghiz that, in contrast to an aristocracy,'. . . the poor do not know, apart from the names of their direct ancestors and their clan and tribe, anything about the distant branches of kinship.' Levshin (1832, m : l l ) also made this point some fifty years earlier. Paraphrasing the aphorism of a famous writer slightly, we can say of nomads: he who controls the present controls the past, he who controls the past controls the future. At the same time genealogies ideologically enable many groups of nomads smoothly to incorporate and adopt outside groups into their own ranks, without making any essential structural changes. It could be supposed that genealogical amnesia (not to be confused with structural amnesia described by Gulliver - see Gulliver, 1955:113ff.), which has been recorded amongst many nomads, is partly to be explained by the fact that it facilitates genealogical assimilation. (On the question of genealogical amnesia and genealogical assimilation amongst the Turkmen see Markov, 1976:218; amongst the Bedouin of Arabia - Rosenfeld, 1965:182; amongst the Bedouin of Cyrenaica - Evans-Pritchard, 1949:56; Peters, 1960:41-2; amongst the Humr Baggara - Cunnison, 1966:6, 113n.6; amongst the Afghans - Ferdinand, 1969:137.) Amongst the northern Somali genealogies are constant and unchanging. Thus when two groups unite in the face of a common enemy they consolidate their union with a formal agreement, not by making up a fictitious genealogy (Lewis, 1962:40). However, in the opinion of Marx (1977:354), more or less detailed knowledge of genealogies is linked to control over a given territory and the ability to defend it. Wherever control is lost genealogists are unnecessary. It is in just this sort of approach that the merging of two planes of social organization, the genealogical and the political, is displayed. Many nomads have controlled their territory successfully without any detailed knowledge of genealogies. In the aspect noted by Marx genealogies are primarily important as a structural principle in the forming of corporative ownershipgroups and in the splitting up of the ownership of key resources. However, these groups can be formed on a basis other than the genealogical one. Amongst the majority of nomads the genealogical principle of descent is linked with unilineality, although in this respect also there are specific variations. Of course, the overwhelming majority of them are patrilineal, a fact which is very consistently emphasized in many native models. However, in practice amongst many nomads, particularly on the lower levels of the organization of economic and sociopolitical planes, links with affines and cognates, establishing supplementary lines of defence, are also very important (Peters, 1960:44-5; Peters, 1967:272-4; Hart, 1973:28-30; Barth, 1973a: 13f.; Digard, 1973:49; Stewart, 1973a:384; Marx, 1978:61-3). Moreover, neither unilineality in general, nor patrilineality in particular 143

Preconditions

of the relations between nomads and the outside

world

are indispensable principles in the social organization of nomads. The Lapps, Koriaks and Chukchi are non-unilineal, i.e. the descent principle of affiliation does not exist amongst them; it is curious that neither do they have extended genealogies. This negative coincidence is obviously not accidental. Certain groups of the Tuareg are matrilineal, or still were so recently, although not completely (Nicolaisen, 1963:137,139,144-6,476-8, 485-6). The Tuareg are not the only exception. According to Nicolaisen (1963:496), the Beja changed to patrilineality only after their conversion to Islam. There are serious grounds for thinking that matrilineality was also preserved for a long time amongst ancient nomads in the Eurasian steppes, the Sarmatians (Khazanov, 1970). Certain tribes of nomads in the Near East and the Eurasian steppes trace their descent from a female ancestor, or else they have female eponyms in their genealogical systems. Several hypotheses have been put forward to explain this (see, for example, Bacon, 1958:178; Peters, 1960:29; Marx, 1977; Marx, 1978), but not one of them can be regarded as conclusive. At the same time, of course, there are no real reasons why we should share the view stated long ago by Robertson Smith (1903:31; cf. Ashkenazi, 1946-9:670-1; Tolstov, 1948:325ff.) that a female eponym is a survival from a hypothetical matriarchy. Segmentary systems in nomadic societies A great deal of writing has been done on the system of segmentary lineages in general and amongst both pastoralists and, particularly, nomads; however, none have fully clarified the picture. The discussion about the extent to which a system is an ideal model and the extent to which it is reality still continues (see, for example, Evans-Pritchard, 1940:141-2; Fortes and Evans-Pritchard, 1940:5f.; Bacon, 1958; Sahlins, 1961; Sahlins, 1968:50-2; Smith, 1956; Bessac, 1965; Gellner, 1969:36f.; Gellner, 1973:3-6; BlackMichaud, 1975:63, 87; Marx, 1977:359; Digard, 1978:315-17; Service, 1975:65f.; Salzman, 1978; Tapper, 1979:48; Asad, 1979:421-2). In my view part of the reason why the discussion continues is contained in the fact that there are different interpretations of the system of segmentary lineages itself. One of the most detailed definitions is provided by Sahlins (1961:330) who singles o u t ' . . . six salient elements of segmentary lineage organization: lineality, segmentation, local-genealogical segmentation, segmentary sociability, complementary opposition (or the massing effect), and structural relativity.' Sahlins notes that where the elements which have been singled out are not to be found in their entirety, the segmentary systems belong to other types. Nevertheless, the social organization of many societies is frequently characterized as a system of segmentary lineages on the basis of a more limited number of characteristics. 144

Segmentary

systems in nomadic

societies

In this respect an instructive example is provided by Middleton and Tait (1958:1-31) who have singled out several variants of such a system in their African material. But when they attempted to specify the common features in all these variants all they found was that: 'In all of them the segmentary principle is operative and relations between territorial groups are conceived in terms of descent, in terms of lineage or clan system' (1958:29). In such a definition, it is essentially only the principle itself, usually the genealogical one, of segmentation which is preserved. And, indeed, this principle has been observed amongst the majority of nomads, both in the present and in the past (Gellner, 1969:48-9). Segmentation, however, does not necessarily signify internal uniformity amongst the segments themselves (see, for example, Nicolaisen, 1963:144-6 on the Tuareg of Ahaggar); and the structural uniformity (structural relativity) of segments from different levels should not disguise the distinction (at least a partial distinction) between the functions which they fulfil (Gellner, 1969:48-9). I shall now consider segmentary systems, but only from the sociopolitical aspect, although, of course, they deserve to be examined from other viewpoints as well. In this respect an interesting interpretation of segmentary systems is put forward by Gellner (1969:51; cf. 1973:4), following Evans-Pritchard (1940; 1949). The segmentary system is characterized as 'stability-withoutgovernment', i.e. primarily as a way of employing a specific mechanism to maintain internal order in conditions where social differentiation has not developed sufficiently. It scarcely need be doubted that this kind of system does, indeed, exist. Gellner has described it in detail, using the mobile pastoralists of the High Atlas as his example. The question is how often does it exist and then, where it does exist, is it only this mechanism which maintains internal order or is it supplemented by others? It would appear that asymmetric segmentary systems, in which structural relativity and balanced opposition are upset, are most characteristic of the types of nomadism which are to be found in the Eurasian steppes, the Near and the Middle East. Thus, although the organization of the Bedouin of Cyrenaica is sometimes regarded as typical of a system of segmentary lineages (Evans-Pritchard, 1949:54-5), Peters (1967:271) writes that there is no balanced opposition in it, for segments of one level differ amongst themselves, both in size and strength. Even where balanced opposition does exist, in practice it is frequently upset (see, for example, Barth, 1973:13). But the main point is that in the majority of systems in the types of nomadism described in the preceding paragraph, the reciprocal position and interrelations between the segments of one level are not identical, not only from the point of view of corresponding numbers, size and strength, but also 145

Preconditions

of the relations between nomads and the outside

world

of political weight. Whilst the majority of segments of one level are thought to be more or less equal, one of them, from the point of view of descent, notoriety and, most importantly, political position stands higher than the rest, 'is more equal than others'. For example, amongst the Bedouin of Arabia '. . . the office of section-chief is vested in chiefly lineage as distinct from non-chiefly lineages' (Sweet, 1965:143). Amongst the Shah Nawazi Baluch '. . . there is a chiefly lineage which provides the tribal Sardar' (Salzman, 1971:186). Amongst the Kababish '. . . virtually all political power and privilege was held by the chief's lineage' (Asad, 1970:XV). Middieton and Tait (1958:14) suggest that the similar systems which have been discovered in Africa are a variant of the system of segmentary lineages. But the fact that there are dominant segments and leaders, which are the natural consequence of such segments, contradicts both the structural principle and the supposed egalitarian essence of the given system. Evans-Pritchard spoke his mind quite unequivocally on this point both in 1940 and in 1949. Compare: '. . . there is no association, class or segment which has a dominant place in the political structure through the command of greater organized force than is at the disposal of any of its cougeners' (Fortes and Evans-Pritchard, 1940:14) and: 'The tribal system, typical of segmentary structures everywhere, is a system of balanced opposition between tribes and tribal sections from the largest to the smallest divisions, and there cannot be any single authority in a tribe' (Evans-Pritchard, 1949:142). Fernea (1970:105) is quite right when he writes: 'The existence of a dominant lineage group, a group having proportionally greater control of resources and ultimately greater power within a tribal group than any other segment, runs counter to the classic model of the segmentary system . . . for it is in a sense the end result of conditions undermining those checks and balances which keep structurally equivalent segments equal in politicoeconomic terms' (cf. Smith, 1956:53, 57; Spooner, 1973:35). Whilst I do not deny that there is, indeed, segmentation in systems with politically dominant segments, I look upon the latter as a distinct variant and call them differentiated segmentary systems. In the appropriate conditions it can happen that, from the structural point of view, the ruling segment ceases to be a replica of the other segments, fully or partially falls away from the genealogical cliche of the given society, acquires distinct laws of kinship and descent and, most importantly, turns into a distinct estate. Examples of this are the estates of Ashina amongst the ancient Turks and of 'white bone' amongst the Mongols and Kazakhs who traced their descent from Jenghiz Khan. These systems may be called stratified segmentary systems. At this point I 146

Segmentary

systems in nomadic

societies

should like to emphasize that the notion of stratified segmentary systems is narrower than is that of social stratification. If a system is segmentedstratified, then the society as a whole is stratified. But if a society is stratified, its segmentary system is not necessarily stratified. It is just that stratification in this latter case develops along different lines. However, it should be borne in mind that all variants of segmentary systems are inter-reversible and, in the right conditions, are able to transform themselves, thus demonstrating that social processes in nomadic societies are reversible. In order to establish more exactly at this point my own view of nomadic segmentary systems, I must move ahead a little and anticipate some of the conclusions of later sections of the book. In my opinion social differentiation amongst nomads in general and differentiated and, particularly, stratified segmentary systems mainly develop as a result of specific relations between nomads and the outside world. Such relations never cease to operate, but specific things about them are subject to historical changes, although in the past they have basically operated in repetitive patterns. Such changes also affect segmentary systems, which then themselves change. Thus it is possible for situations to arise in which the degree of egalitarianism or, on the contrary, of inequality characteristic of properly functioning segmentary systems do not coincide or, more often, do not fully coincide with declared ideologies (and, particularly, the sub-ideologies of the different groups and strata of the given society). Ideologies may be more or, on the contrary, less inegalitarian than social reality at the moment of observation; this is because they include not only the given moment, but also the past and even sometimes an outline of changes in the future. Below I cite examples of such disparities. At the same time it should be remembered that ideologies do not only reflect social practice, they also play a part in the formation of it and the way it changes. A diachronic study of segmentary systems, wherever this is possible, can lead to a more comprehensive understanding of them, just as the discovery of the direction of the predominant tendency - towards stability or change at the moment of observation or description also can. If it is towards change, the direction of the changes should be ascertained. In an historical approach to specifically functioning segmentary systems amongst nomads, their 'egalitarian' and non-egalitarian tendencies are no longer mutually exclusive. In some cases they can witness to the fact that the system, like the society itself, is in a state of change. But in specific situations they can be complementary, more or less balanced and at the same time they can fulfil the function of maintaining internal order. Finally, it should not be forgotten that amongst nomads there are two sides to segmentation. The first is the ideological side, for the native model, in which the descent principle and genealogies are realized, has clearly 147

Preconditions

of the relations between nomads and the outside

world

defined segmentary features; as Salzman (1978c) has convincingly shown '. . . ideologies are often social structures in reserve.' The second is the practical side, for segmentation appears not only in the self-concep­ tualization of the social organization of nomads, but also in the realization and transformation of this organization.

The upper levels of sociopolitical organization in nomadic societies Coon (1976:319) has compared the largest associations of nomads, tribes and confederations, with 'skeleton organizations, to be used if needed.' This comparison must be amended on just one essential account. A skeleton is invariable, but the upper segments of the sociopolitical organization of nomads are not, nor, in some senses, are the other segments which have their own dispositional side. They emerge in accordance with necessity and disappear, simplify or reform when the need for them diminishes. Their significance increases as the significance of the political functions they perform increases, and then decreases when the need for them declines. In pastoral nomadic societies the production cycle is already guaranteed in the lower levels of social organization. AH that is needed for the optimum pasturing of livestock, utilization of natural resources and necessary mutual aid and co-operation are individual households and the different orders of communities. Such, near enough, is the organization of the reindeerherders of the North; here there are no descent groups, or at best they are expressed relatively vaguely, as amongst the Nentsy. The higher levels of organization amongst nomads do not result directly from productive processes. They are linked not with economics, but with other planes, notably sociopolitical ones, though they can also fulfil distinct and quite important economic functions. I would like to stress once again that although, in theory, we can separate out the different planes of functioning of nomad societal organization, in practice the same structural unit can simultaneously perform a variety of functions. These higher levels of nomadic organization in the first instance may be a response to social requirements; but then their functional role is relatively limited, such as in certain East African pastoral societies. On the other hand, the significance of these higher levels inevitably increases as nomadic societies become more politically integrated. These tendencies may be linked with various factors. Amongst the most important are: the need to allocate rationally key resources; the establishing and regularizing of routes of pastoral migrations; need of a certain order, necessary for productive processes; need for defence; the struggle for livestock, pastures and arable lands; migrations and wars; the desire of 148

The upper levels of sociopolitical

organization 8

certain groups of nomads to subdue others; particularities of relations and interaction with the outside sedentary world, etc. Very tentatively these factors may be divided into three groups: those which operate within one nomadic unit; those which operate between different nomadic units; and those which operate between a nomadic unit, on the one hand, and the outside sedentary agricultural-urban world, on the other. Moreover, their significance and the form in which they appear in a specific historical situation and in different ecological conditions are not identical. The correlations between the listed factors and the relative significance of each of them determine the specific character, stability and level of the political integration of nomads; they also determine the degree of institutionalization and centralization of power, including the strength of the latter, amongst nomads. That same genealogical principle of descent which acts as a structuring agent in many nomadic societies is also operative in the upper levels of their organization. Furthermore, it is possible that it was the need (potential or periodic included) in the upper levels for this which favoured the development of genealogy and, in particular, unilineality amongst nomads and a corresponding plan in their social organization. Perhaps Bessac (1965:392) is partly right when he remarks: 'Given fragmentation of the primary herding unit due to increases and decreases in yield, there is no economic reason why the society should develop a unilineal bias. On the contrary, affiliation should be bilateral and largely associational.' North Eurasian and, partly even, East African nomadic societies serve to confirm this idea. Amongst the former there are no genealogies and unilineality is not widespread; amongst the latter genealogies are relatively poorly developed. The works of Kroeber, Manners, Colson and, particularly, Fried (1975; cf. Biebuyck, 1966; Southall, 1970) have shaken the old idea, which dates back to Morgan and his predecessors, that the tribe is the basic cthno-linguistic, territorial and political unit of primitive societies. Follow­ ing on from them Emmanuel Marx (1977, 1978, 1979) has stated that the problem of the essence of nomadic tribes also needs to be reconsidered. In his opinion the nomadic confederation and partly also the nomadic tribe are primarily territorial units, not military-political ones with the corresponding leadership. It seems to me that the problem is much more complex. A nomadic tribe 11

11 is possible that in this context the notion of 'chain reaction', introduced earlier as an explanation for certain particularities of the way in which pastoral nomadism itself spread is also applicable. Different forms of social organization in o n e set of nomadic societies may have been copied (with different variations) by other societies, for example, in the interests of KdMcl'cnce, etc.

Preconditions

of the relations between nomads and the outside

world

does occupy, of course, a specific territory and in this sense can indeed be regarded as a territorial unit. However, it bears no resemblance to any canton or district with historically or administratively established bound­ aries. Firstly, as I have already pointed out, straightforward territorial ties are not usually to be found amongst nomads, rather they are mediated by kinship and descent. Secondly, in the overwhelming majority of cases a nomadic tribe is also a social unit, if only for the reason that it signifies the upper boundaries of a discrete social organization. These boundaries are not necessarily, of course, stable and unchanging, and the discrete character of the tribe, in common with any other form of social organization amongst nomads, is relative (this point has been dealt with above). Finally, the tribe as territorial unit cannot be separated from corresponding relations concerning ownership of key resources and capacity to defend these ownership rights. (Marx has also, incidentally, written about this 1978:49-50.) Even if a tribe does not act as the direct holder or regulator of these rights, it should still be regarded as the supreme owner and/or guarantor of ownership, at least as the embodiment of such. After all a nomad possesses rights to natural resources not simply because he lives in a given territory, but because he belongs to a given tribe and to a corresponding subdivision of it. 'Descent is the allocation on the basis of genealogical position of rights held by a restricted group as against the world' (Barth, 1966:24). Nomads of the Near East in the present and recent past form the basis of the material which Marx uses to corroborate his opinion. But the Bedouin of the Negev and Sinai, subjugated by powerful centralized states, are hardly very representative examples on which to base such far-reaching conclu­ sions. The reports of early observers, according to which it was the tribe which was the supreme owner of grazing territory (amongst the Bedouin of Arabia, for example), should not be dismissed out of hand. By way of example I quote Tamisier (1840:256), who stated unequivocally: 'Each portion of territory belongs to a tribe: the tribe can dispose of this territory as it wishes: it is a property its fathers have handed down along with their blood.' (See also Wallin, 1854:122; Burton, 1893:113-14; Jaussen, 1908:238f.; Dickson, 1951:582). Admittedly, this alone was not always sufficient guarantee of stable institutionalization and, even more so, of centralization of power. Thus, not all tribes in Arabia had a supreme chief (Marx, 1977:348-9); however, judging from the innumerable descriptions of travellers (for a summary of these see Pershits, 1961:153ff.), the ones who did not have such a chief were decidedly in the minority. At any rate, amongst the Bedouin of Arabia, there were no 'tribes without rulers'. Even in those tribes in which there were no supreme chiefs, subdivisions of a lower taxonomic level which had become part of the tribe had chiefs unci leaders. 150

The upper levels of sociopolitical

organization

My conclusion from this is that a nomadic tribe is never a purely territorial unit (cf. Salzman, 1979a). First and foremost it is a political organization. Incidentally, even in the last century there were scholars who believed that tribes of the Bedouin were not associations of people related by blood, but military-political units (see, for example, Jacob, 1897:222). Wherever territorial ties are not supplemented or mediated by political ones (as used to be the case amongst the Koriaks and Chukchi) what we have is an ethno-linguistic unity. Whether such a unity can be regarded as a tribe is a question open for further discussion. Just as political organization should not be identified with government, the presence of political functions in a tribe should not be determined only by the existence in that tribe of a strong institutionalized and, even more so, centralized power (cf. Marx, 1978:50). The two cannot be completely separated from one another, nor can they be identified. Even the degree of social differentiation in a given nomadic society and the degree of institutionalization and centralization of power in that society do not necessarily correlate completely. A diffusely spread type of political power is a peculiarity of many nomadic units; only in specific conditions are there centralized forms of political power. But an adequate impression of nomadic tribes is scarcely to be gained only through examining them from within. Nomadic societies are never closed societies. The political functions of a tribe are most clearly manifested when that tribe confronts the outside world, primarily the sedentary world, but partly also the nomadic one. It is no coincidence that the majority of factors behind political integration, also concentration and centralization of power, amongst nomads are connected with their relations with other sedentary and nomadic societies. Even such factors as the regulation of routes of pastoral migrations, which conventionally are regarded as internal ones, frequently depend on the position of the given nomadic society in the outside world (as, for example, amongst many nomads in the Middle East). Gellner (1969:2-3) has suggested that there is a distinction between 'primitive' and 'marginal' tribalism. In his view the latter emerges on the borders of agrarian-urban states and reacts against the power of those states. Fried (1975) completely dismisses the notion of 'primitive tribalism'. In his view tribalism is always secondary and emerges only under pressure from more powerful societies. It is true that in the majority of cases nomadic tribalism is secondary and marginal for, after all, relations with the outside world are one of the first reasons for the emergence of tribalism. Such relations are understood here in the widest sense to include both defensive and aggressive relations, and relations both with agricultural and urban societies (in the first instance they are with such societies) and sometimes with other nomadic societies. This does not mean that internal factors contributing to the emergence of political tribalism amongst nomads should • SI

Preconditions

of the relations between nomads and the outside

world

be ignored; it is just that in the majority of cases they played a secondary and attendant role (cf. Pershits, 1961; Barth, 1962:347; Spooner, 1973:35 on the nomads of the Near and Middle East). However, the political tribalism of nomads is internally contradictory. To some extent it imposes some restrictions on the freedom of choice of the lower productive sub-units. At the same time the specific character of production amongst nomads is such that the individual household and community need to have sufficient freedom of action and to be able to make independent decisions. This contradiction to a great extent explains both the periodic emergence of nomadic polities and their instability and impermanence. Even more does it explain why it is that strength and centralization of power in a nomadic society frequently lag behind the degree of social differentiation in that society. It is the specific situation which determines which forces are to become predominant - centripetal or centrifugal ones, and which is to be the predominant tendency - towards diffusion of political power or towards concentration of this power. Again, it is the specific situation that causes the emergence of political tribalism which determines the size of the nomadic units. If the size of the units exceeds the size required, fission is inevitable (see, for example, Awad, 1962:333 on the nomads of the Near and Middle East). In conclusion I should like to say something about those nomadic units or associations which are frequently referred to as 'confederations'. In my opinion this term is very inappropriate, because nomadic associations are not always formed on a voluntary basis. In all circumstances they emerge for military-political reasons (see, for example, Marcais, 1913:242ff. on 'confederations' amongst nomads in North Africa in the Middle Ages) and therefore, despite the fact that sometimes they do not have a com­ mon leadership and clashes within them of the different tribes cannot be entirely avoided, they can and should be regarded primarily as political units. Property-inequality and social differentiation in nomadic societies Looking at nomadic societies from the inside, two potential sources of social differentiation may be singled out: private ownership of livestock and positions of leadership in the social organization. Private ownership of livestock, differences in the domestic cycle and the inherent instability of pastoral nomadism inevitably lead to property9

9

By social differentiation I understand any forms of social inequality and stratification (status, rank, estate, etc.); by social stratification, the presence in a society of hierarchical hereditary strata and groups with different rights and duties, and which occupy different positions in public life, the running of their society, and/or in production and/or distribution.

152

Property-inequality

and social

differentiation

inequality. This inequality means some individuals and their families have more livestock than others (cf. Bonte, 1977). In common with the peasant community, the nomadic community is interested in property-equality amongst its members. Where inequality is too pronounced the system of co-operation and mutual aid which exists in the community is put in jeopardy. Sharp decline in the material circumstances of individual households makes such households unsatisfac­ tory partners in the community. In such cases impoverished households either fall away from the community or become dependent on richer households. In both the nomadic and the peasant community there are always two simultaneously active tendencies: towards levelling and the ironing out of internal differences, and towards furthering property-inequality and increasing spheres of activity in which relations of dependence and exploitation between different households operate. The realization of both tendencies goes beyond the confines of the community and depends on the situation in the society as a whole; where nomads are concerned it frequently can even depend on their relations with the outside world. Nomads in need of a work force can rely on the poor members of their community to pasture livestock and do other work in the households of their richer relatives and neighbours. However, as a rule, in the nomadic community close relatives are not exposed to exploitation as this would jeopardize essential social ties. Nevertheless, even in the nuclear community there are ways in which rich pastoralists can use outside labour and benefit from doing so. Sometimes this is achieved by joining poor families unrelated by kinship to the prosperous kin core of the community. For example, migrants from other tribes lived in every nomadic camp of the Rwala; these migrants were called kusara and were guaranteed the patronage of certain members of the Rwala (Musil, 1928:267-8; cf. Peters, 1960:43 on the Bedouin of Cyrenaica). Amongst the Kazakhs the core of an aul community consisted of families of close agnates, and sometimes also of affines, but emancipated slaves and konsy (kongsy) were also allowed in. The latter constituted a distinct social category which was made up primarily of poor individuals who had lost their links with their own auls and were living in alien auls (Viatkin, 1947:132n.l; Sabyrkhanov, 1969:150n.28). In Kirghiz auls there also lived outsiders who were the descendants of slaves (Iliasov, 1963:345). Amongst the Nentsy the well-off reindeer-herders joined up with the weaker ones in the parma and, according to an established tradition, took an equal turn in the herding of the stock, despite the fact that one herder might own 1,500 head and another 30 head (Brodnev, 1959:77). An analogous institution, 'the permanent herding association . . . based on shared food and labour' exists amongst the Karimojong. These institutions emerge in 153

Preconditions

of the relations between nomads and the outside

world

situations such as when one family owns a large number of animals, but has insufficient labour in its ranks, and when another owns few animals, but has an excess of labour (R. and N. Dyson-Hudson, 1970:110; cf. Stenning, 1959:53 on the Fulani). However, the opposite tendency must also be taken into account. A number of circumstances impede and prevent property-inequality becoming the sound and stable basis of social differentiation amongst nomads. To begin with, that same economic instability which is characteristic of nomadic society as a whole is also to be found on the individual family level. Reallocation of property, primarily of livestock and pastoral products, can promote social advancement and, to some extent even, disguised exploitation. At the same time it is connected with the needs of social integration and is also, to some extent, a particular form of economic insurance. Such reallocation usually takes the form of reciprocity and redistribution in their various specific manifestations. The correlation of reciprocity and redistribution is different in different nomadic societies. All that can be said is that the less social differentiation there is in a given society, the greater is the importance of balanced and non-balanced reciprocity in that society. Non-balanced reciprocity does not require the taker to return the equivalent of what he has taken, but it does place moral obligations on him vis a vis the giver and increases the prestige of the latter. Balanced reciprocity requires that whatever is given by some means or other is returned, although not necessarily in exactly equal proportion (cf. Sahlins, 1974:193-5). Redistribution naturally is more connected with social inequality. In so far as they also fulfil the functions of collective mutual aid and insurance, redistributional institutions can be at the same time the economic channels of social differentiation. But as is frequently the case in societies other than nomadic ones, redistribution does not so much do away with and limit reciprocity; rather it overlaps with the latter which continues to exist even in nomadic states. Balanced reciprocity amongst the Tatars of the Golden Horde in the fourteenth century is described by al-'Umari. 'When an animal belonging to one of them begins to weaken, for example, a horse, or cow, or sheep, he will kill it and together with the other members of his household he will eat a part of it, and [part] give to his neighbours, and when one of their sheep, or cows, or horses weakens, then they will kill it and give [part of it] to those who had given to them. For this reason in their houses [never] is there a shortage of meat' (quoted in Tizenghauzen, 1884:23-31). A similar custom existed amongst the Bedouin (Doughty, 1881, i:442-3). Amongst the nomads of Arabia help was proffered to individuals at the time of key events in their life-cycle (marriage, circumcision, etc.), in extenuating circumstances (payment of composition for murder, demands of hospitality) 154

Property-inequality

and social

differentiation

and also in the event of loss of livestock, for example as the result of disease (Jaussen, 1908:273; Musil, 1927:461f.; Pershits, 1961:76). According to Doughty (1888, i:344), robbery was regarded as a misfortune to be shared by all, thus a sheikh would insist that all the members of his tribe helped to replace within one or two days the livestock of those who had been robbed (cf. Musil, 1928:452 on the impoverished Rwala). Another apposite example is saun (soan) (cf. wadiya amongst the Bedouin of Arabia), a Kazakh word which in Soviet anthropology designates the various forms and terms by which rich stockowners entrust the pasturing of their animals to the poorer members of their society. (For a description of those amongst the Kazakhs see Georgi, 1777, i:131; amongst the nomads of the Near and Middle East - Doughty, 1888, n:140; Jaussen, 1908:273; Pershits, 1961:124-6; Swidler, 1973:28-9; amongst the Fulani Stenning, 1959:41, 62). For a long time many Soviet anthropologists unreservedly regarded saun as a means of exploitation, frequently even a feudal one (Tolstov, 1934; Viatkin, 1947:119-21; Tolybekov, 1971:159ff.). However, Pershits (1973) has shown that saun includes various forms of redistribution of stock, purely reciprocal forms amongst them. Dahl and Hjort (1979:21) single out two forms of stock-association amongst African pastoralists; the first ('stock-alliance') is based on a system of gifts, the second ('stock-patronage') on a system of loans. But even in those forms of saun in which there are several elements of exploitation or in which, more often than not, social inequality is promoted, there is another aim. They discourage poor households from falling away from nomadic societies, thus helping to prevent the disintegration of the societies in question. An analogous role, although this is not its only function, is played by stock-patronage in Africa. Moreover, it sometimes happens that with time borrowed animals come to be regarded as the property of the 'client' or of his descendants. In this way this form of redistribution of livestock can serve to equalize property differences (Dahl and Hjort, 1979:23). It is noteworthy that the strongest and most clear cut exploitative features of the wadiya were manifested not within a given nomadic group, but in the relations of different groups when rich and noble Bedouin left the pasturing of their small stock to sheep-herding tribes or sheep-herding subdivisions of their own tribes (Philby, 1922, i:59; Muller, 1931:50f.; Dickson, 1951:109-10; Pershits, 1961:125-6). Irons (1979:365 , 367) has rightly pointed out that amongst nomads redistribution has not been as widespread and significant as in other societies with different economic systems. Evidently, this is primarily to be explained by the fact that centralized accumulation of livestock, even if justified as a necessity for insurance purposes, is practically impossible. Relations of the saun type amongst nomads partly serve as an alternative to redistribution. 155

Preconditions

of the relations between nomads and the outside

world

However, redistribution may be observed clearly in many nomadic societies. In Oghuz epic literature it is said of the aristocracy: 'Without ruining himself with property a man cannot glorify himself with generosity' (Kniga moego deda Korkuta, 1962:11). Amongst the Bedouin of Arabia it was accepted that sheikhs from time to time organized public feasts and gave presents to the members of their groups with few possessions (Dickson, 1951:53). Cases are known of Bedouin sheikhs dividing their revenues 'among the mass of their people' (Burckhardt, 1831:308, cf. 331). Amongst the Kazakhs bais, biis and sultans (i.e. members of the ruling and well-to-do strata) often organized lavish feasts, following the saying: 'The dog that is hit with a fatty bone never whines' (Tolybekov, 1959:95). Other redistributive mechanisms which operate between different subdivisions and groups of nomads must also be taken into account. Whether or not Sweet (1965:142ff.; cf. Pershits, 1961:131-2; Spooner, 1973:11) is right in suggesting that Bedouin ghazw, livestock raids that is, are an institutionalized form of stock-redistribution; in practice, they do partly lead to this. An analogous institution for which Soviet anthropologists use the Kazakh word barymta used to be widespread amongst nomads in the Eurasian steppes. The Bedouin ghazw and Kazakh barymta were in many respects different from, for example, Turkmenian alamans. The latter took the opposite form of a protected, limited and random plunder on sedentary, basically Persian, peoples who were practically unable to answer back. One of their basic aims was the seizure of slaves. Ghazw and barymta, when operating between pure nomads, were a two-sided affair and led to an uneconomic circulation of livestock between different nomadic units and their subdivisions, although not without losses. It is curious that amongst the Kazakhs in the nineteenth century barymta was sometimes used by poor nomads as a way of seizing stock from the Kazakh nobility. One more factor hindering the unlimited development of propertyinequality amongst nomads must be mentioned. Although generalizations about the minimum number of livestock necessary for the running of an independent household are practically impossible (see pp. 29-32), there is such a number for each specific society at any specific time. Thus, impoverished nomads unable to procure a sufficient number of livestock by way of the institutions mentioned above have two main alternatives - either they can enter the service of rich stockowners, or they can supplement pastoralism with other means of subsistence, which often entails partial or complete sedentarization (Barth, 1964:108f.; Barth, 1964a:75; Dahl and Hjort, 1979:27). However, few stockowners are so rich that they are 10

10

Levshin (1832, m:85) observed that barymta decreased the number of stock amongst the Kazakhs, for stock tended to perish when driven off at the greatest possible speed.

156

Property-inequality

and social

differentiation

incapable of dealing with their own livestock by themselves or with the help of mechanisms of co-operation and mutual aid (including saun) in their own society; at any rate such rich stockowners are a great deal less common than poor ones. Thus, a certain number of poor stockowners are all the time being excluded from nomadic society (see pp. 82-4), and this also to a certain degree is able to limit property inequalities there. Zhitetsky (1892:25) wrote of the Kalmucks that, '. . . relatively well-to-do families roam in the steppe, and those of them which become poor and lose their stock year in, year out are forced out by the steppe to beyond its boundaries. Of the families which own little stock only those remain in the steppe, and there are relatively few such families, which are attached to the khuduks [draw-wells] of rich pastoralists and to khuruls [Buddhist monasteries in Kalmuckia] as shepherds or workers.' In this way, the desire for unlimited increase in numbers of livestock (which is, in fact, impossible because of the reasons which were examined in Chapter 1) does not necessarily mean that property-inequalities are constantly increasing. Demands of co-operation and mutual aid, the necessity for which is dictated both by productive and social needs, favour the development of a more or less clearly defined tendency towards limiting differences if not in the ownership of livestock, then in the way in which animals and their products are utilized; otherwise nomadic society is in trouble. Property-differences are often only temporary. Nevertheless such dif­ ferences still can encourage social mobility, the attaining of high status or rank and thus a certain degree of social differentiation; but in themselves they are rarely capable of turning the latter into stable and hereditary social stratification. Amongst East African pastoralists, for example, the position of creditor in stock-patronage '. . . enables a man to build himself into a position of bigmanship' (Dahl and Hjort, 1979:23), but no more. Property differences do not so much create as strengthen hereditary social inequality wherever it exists. It would be too simple, but incorrect, to follow Lattimore (1979:483) and assert that: 'wealth tended to become hereditary, so that a tribal aristocracy emerged.' Of course, a rich stockowner could gather a certain number of people around himself, form a bodyguard out of them and with its help strengthen and increase his own influence and power. Thus did the brave and generous sheikh gather together followers, some of them even from other tribes (Oppenheim, 1900, n:88-9). However, such people were tied to their leader only by expediency and personal loyalty and they had to be maintained; stock had to be expended in order to maintain them. Their labour was used in the household, but not very much because primarily they were warriors, ami also because ways in which such labour can be utilized in a pastoral economy are limited. Social gains sometimes turned into property losses. IS7

Preconditions

of the relations between nomads and the outside

world

One way of avoiding such losses consisted in seizing stock and other valuables from other societies or even from other subdivisions of one's own society. It is no coincidence that heightened military activity, internecine wars included, is usually characteristic of nomadic units in which such a social stratum as the Mongol noktir or Kazakh tiilenguts have been re­ corded. Vladimirtsov (1934:80f., 95) equated the twelfth-century Mongol khan with a leader of a robber band whose behaviour was determined by his need to maintain an armed force, his nokod. It was on this account that neighbours were raided for booty. Amongst the Kazakhs the institution of tiilenguts, came to nothing in the nineteenth century when Russian power enforced internal peace amongst them (Bekmakhanov, 1947:48ff.; Bizhanov, 1969:169). However, such military activity may be the result of various factors. Sometimes it can itself stimulate the emergence of military leaders with surrounding bands of men for whom raids and war become a way of life. In this case causes and consequences are extremely conditional and can vary from place to place. To my mind, what it is most important to emphasize is that in all circumstances this method of socially implementing propertydifferences is linked to external factors. Property-inequality amongst nomads frequently acquires particularly marked dimensions in two cases - where there is a nomadic state, or where nomads are incorporated into a sedentary state. In such circumstances the state is the guarantor of property and at the same time weakens redistributive mechanisms. Barth (1966:16-17) has cited one instance of transforma­ tion of a traditional society (although not a nomadic one) in contemporary conditions; he has observed that in Swat redistribution occurs more often when valuables are hard to sell in the market, and quickly diminishes when things in the market improve. Dahl and Hjort (1979) have shown that a similar picture may be observed amongst East African pastoralists. Admittedly, many scholars (Lambton, 1953:289; Barth, 1964:103f.; Stauffer, 1965:294; Ferdinand, 1969:142ff.; Johnson, 1969:77) cite ex­ amples of how in the Middle East stockowners convert livestock into capital in the form of land and then frequently adopt a sedentary way of life. However, this phenomenon is a recent one; it is local and limited, and is characteristic of the present period. It signifies a partial disintegration in corresponding nomadic societies. Thus, in Baluchistan, especially in Kalat, only very recently did land become something which was bought and sold (Swidler, 1973:29). Incidentally, even in the present day sometimes the reverse side of the picture, the conversion of capital into livestock, may be observed. The Somali sometimes spend many years working outside their country, for example, as sailors, so as to then return home, procure livestock and once again lead a nomadic life (Lewis, 1961:32). 158

Property-inequality

and social

differentiation

In ancient times and the Middle Ages there was no land-ownership in the capitalist sense. Acquisition and preservation of land by nomads by way of monetary-market transactions, if it had spread at all, had done so only to a very, very limited extent. Nevertheless, it is true that control and ownership of non-pastoral production and the means of production, particularly of cultivated lands, reenforce social differentiation in nomadic society and the power of leaders over ordinary nomads (Capot-Rey, 1962:304; Peters, 1967:176; Salzman, 1978:132,135; Salzman, 1979:439). Thus, when conditions were favourable a nomadic aristocracy would strive to acquire cultivated lands, if only because corporative ownership of key resources restrains processes of social differentiation amongst nomads. Although they disliked involvement with agriculture, nomads would willingly become landowners. But such cultivated lands were not capital and primarily they were acquired by way of conquest, seizure and rewards. One other source of social inequality in nomadic societies must be mentioned; this is the frequent presence within them of slaves and other dependent and exploited groups and individuals. They encourage inequality in that wherever there are dependants there must exist those who are depended upon, wherever there are dominating individuals the dominated are to be found. Thus, on the strength of its role as manager of social affairs and leader of society, a nomadic aristocracy derives the greatest benefit, both from the property and social points of view, from the presence of these dependent groups. Members of the aristocracy act as their protectors, patrons, etc. This, in turn, gives extra weight to the aristocracy in its dealings with the ordinary, free members of their society which can work against traditional institutions for limiting social inequality. Nevertheless, the presence of such groups in nomadic societies, as a rule, is insufficient to make the society stratified. One reason for this is that nomadic economy is not sufficiently diversified. When dependent groups and individuals are involved in the same pastoral production as other members of society, there must be relatively few of them for their dependent position to be preserved. If this does not happen then practically, if not legally, their status gradually becomes the same as that of the ordinary, free members of the society. One of the many examples which could be quoted here is the otegii-boyol in Mongolia in the twelfth century and the beginning of the thirteenth, literally, according to Pelliot, 'ancestral slaves', or in his translation 'hereditary serfs' (Pelliot and Hambis, 1951:85-6). The essence of this social group has been a subject of discussion in Mongolian studies for many years. Vladimirtsov (1934:65) looked upon unagan bogol (i.e. otegu-boyol) as'. . . feudal serfs who could not freely break their ties with the ruling clan.' However, he greatly overestimated the degree of dependence of otegu11

159

Preconditions

of the relations between nomads and the outside

world

boyol. 'These "slaves" were, however, rather vassals, retaining their freedom of movement and even their tribal organization, and many of them were promoted to high positions' (Pelliot and Hambis, 1951:85-6). Of course, dependent groups are sometimes involved in non-pastoral economic activity. But in this case they occupy a marginal position and are situated on the borders of a nomadic society with a wider social continuum which is connected with the outside world. The ease with which nomads called anyone occupied with agriculture a slave should not lead to misconceptions. In the eyes of nomads an agriculturalist is a slave because he is tied to one place and is enslaved by his own arduous labour, unable to resist them in any proper way. The situation with slaves is roughly similar. Until recently slavery was characteristic of practically all nomads in the Eurasian steppes, and the Near and Middle East; at the same time it was never of great significance from the point of view of production. In this respect only the Tuareg are something of an exception. In my view the main reason for this is that in principle a nomadic pastoral economy cannot use slave labour on a large scale (Khazanov, 1973:415-38; Khazanov, 1975; Khazanov, 1976). Of course, ownership of slaves increased the prestige, influence and power of their owners (see, for example, Stenning, 1959:65 on the Fulani). In this respect perhaps the best example is the use of slaves as bodyguards and in armed forces by the most powerful Bedouin sheikhs, for example, the Shaalans and Rashidids (Musil, 1928:59,277; see also Doughty, 1888, i:553; Oppenheim, 1890, n:89; Pershits, 1961:101; Rosenfeld, 1965:177). It is known that there were black slaves in such bodyguards in pre-Islamic times. However, only a sheikh depending on outside sources of income could maintain such armed forces. The same is the case with bodyguards made up of slaves as it is with those which are made up of free men. The employment of slaves in the pasturing of stock or in domestic work was not usually of any essential economic significance, the more so because a nomadic society is not easily able to prevent slaves escaping. It is no coincidence, therefore, that the majority of slaves or their descendants were set free and partially adopted in the nomadic society; however, their humble status and economic dependence on their former masters continued (see, for example, Wallin, 1850:26 on the nomads of the Near East). 12

13

11

12

13

The case of the Tuareg w h o , at first glance, seem to contradict this assertion will be dealt with below. See pp. 180-1, 278. But in this respect 'uncivilized' nomads were in no way exceptional. In Chinese essays of the T'ang period all non-Chinese were barbarians and when barbarians fought against China they were called slaves. This tradition traces its roots back to ancient times. It is interesting that in antiquity shepherd slaves always lived better than other slaves who were involved in production. The impossibility of day-to-day control, the nature of the work which demanded some initiative, and the necessity for arming shepherds so that they could protect their herds prompted masters to allow their shepherd slaves a certain amount of initiative.

160

Property-inequality

and social

differentiation

Another situation was created when slaves were settled on land and their dependent status preserved. But this already has to do with relations between nomads and agriculturalists and will, therefore, be dealt with below. It also must be pointed out that the presence of dependent groups in nomadic societies cannot necessarily be regarded as a manifestation of their internal social differentiation. This is because the origins of these groups are linked with primarily the relations between a given nomadic society and the outside world, and more precisely with the consequences of those relations. Social inequality in nomadic societies emerges primarily when a distinct political power, with the corresponding specialized functions of leadership and management of society, are needed. The three sources of this power internal organizational-managerial needs, need of interrelations with other nomadic societies and need of interrelations with sedentary agriculturalurban societies - not only encourage the emergence of a ruling stratum, but in the end they determine the social positions of its members and the conditions of its transformation into a hereditary aristocratic estate. The more the need for leadership manifests itself and the more important it is for a society, the more solid and stable is the position of the ruling stratum. It also is important to note that social needs not only create and consolidate the ruling stratum, the ruling stratum also creates corresponding needs. If the latter were not the case, the picture would be too one-sided. It is in its capacity as the ruling stratum that a nomadic aristocracy enjoys specific privileges in its society. Some examples of such privileges are utilization of the best pastures, of voluntary gifts and even of more or less fixed taxes in kind and labour from ordinary nomads. High social position is the most reliable and perhaps the only guarantee of property-differences; furthermore, the traditions of collective social responsibility are sometimes used to secure these differences. Nevertheless, the needs of a society proper are the least likely factor to consolidate social inequality within it. First, this is because they are in themselves the most limited and impermanent needs. The integrating processes in nomadic societies are not sufficiently stable. Centrifugal tendencies counteract centripetal ones. Secondly, they emerge partially under the direct or indirect influence of the outside world, as a result of the functioning of the given nomadic society in a wider economic and sociopolitical continuum. Thirdly, a nomadic society can only exist independently in this continuum if it is to some degree united and consolidated, and if public needs and interests predominate over individual and group ones. This state of affairs acts as a definite check on the 14

Although I examine these three factors separately, I do acknowledge that all three of them can operate at the same time within one nomadic society. Yet another reason why it is hard to separate them in that I lit* iictivution of one of them can activate the others.

161

Preconditions

of the relations between nomads and the outside

world

development of social and property differentiation. In such societies, in both form and essence, the ruling stratum represents public interests. Other factors hindering the development of social differentiation in nomadic societies, namely mobility and low density of population, must also be taken into account; they have already been noted by different scholars (Burnham, 1979:362; Irons, 1979:362; cf. Marx, 1978:52n.l6). However, as a whole they are secondary and surmountable. Therefore, internal processes within nomadic societies which are linked with social differentiation are reversible and, most importantly, not too intensive. In specific, fairly rare cases they can make for a stratified society, but never a state. At any rate I know of no examples of a state emerging as a result of the internal development of a nomadic society. The second source of social differentiation is linked with the interrelations of different nomadic societies, particularly in cases where these relations result in one group of nomads submitting to others. Differentiation must be made between two types of submission. In the first type are those instances, which were briefly examined above, in which within one unit, alongside the dominant groups and subdivisions, there are other groups and subdivisions which do not possess full rights, and which are dependent and exploitable. In the second are those instances in which relations of dependence are established between different units. Both these types of submission were well known throughout the independent history of the nomads of the Eurasian steppes (see, for example, Khazanov, 1975; Khazanov, 1978 on the Scythians; Bichurin, 1950:144; Bichurin, 1950a: 161, 190-1; Taskin, 1973:54 on the Hsiung-nu; Bichurin, 1950a: 186, 229 on the K'ang-kiu (Kanju); Destunis, 1860:420; Gumilev, 1967 on the ancient Turks; Agadzhanov, 1969:159n.3 on the Polovtsians (Qipchaq); Anninsky, 1940:88-9; Polnoe sobranie russkikh letopisei, x, 1885:90 on the Mongols of the Golden Horde). Particularly stable was the type of nomadic unit in which one tribe dominated the others within it and the latters' dependence was of the vassal-tribute type; in the ideology of the dominant tribe the latter were traditionally regarded as slaves (Khazanov, 1975:154f.).is However, in the Eurasian steppes such relations were always less important than the dependence established between nomads and the sedentary population (Khazanov, 1979). The same was true of the Middle East. Nevertheless, in the tribal units of the Afghans one tribe was usually the leader whilst the others were its vassals. Amongst the Lohanis tribes in the sixteenth and seventeenth centuries the leading tribe was the Daulat 15

Of course, this sort of 'slavery' should not be understood in the literal sense of the word. Nevertheless, the references to it in the tradition of sedentary peoples have encouraged certain Soviet scholars to look for slave-owning relations amongst nomads, particularly in antiquity. For a critique of that view see Khazanov, 1973; Khazanov, 1975; Khazanov, 1976,

162

Property-inequality

and social

differentiation

Khel, amongst the Khakhay in the fifteenth and sixteenth centuries it was the Yusufzais, and amongst the Ghoria Khel it was the Khalils (Reisner, 1954:98). The dependency of certain groups of nomads on others was of far greater significance in the Near East, for here it was closely linked to their economic specialization (Rosenfeld, 1965:77). As late as the beginning of the twentieth century almost all the semi-nomadic tribes of Syria paid tribute to the Anaza Bedouins and those of Upper Mesopotamia to the Shammars, those of Northwest Arabia and Hijaz to the Banu Sakhr, Huwaitat and certain others, and those of Northeast Arabia to the Mutair and certain others (Blunt, 1889, i:29-30; Musil, 1928:59-60; Pershits, 1976:295). It was the case in North Africa for most of the Middle Ages that amongst those tribes which paid tribute and were dependent on the makhzan (tribes supplying contingents to the state and exempt from taxation) there were semi-nomads and nomads even. Representatives of the second type of dependency were those individuals and groups who did not enjoy full rights, allies, clients and all those seeking the protection of second degree asabiyya and were linked with the aristocracy by various contractual ties (Marcais, 1913:242; N. A. Ivanov, 1963:192; cf. Stewart, 1975 on Mauritania). By descent they were from different tribes, but were frequently adopted by the dominant tribe (cf. Peters, 1967:173-4 on the possibilities of status-change from client to nobleman in Cyrenaica). However, even this source of social differentiation and political power in nomadic societies is neither stable, nor reliable. From the economic point of view the internal difficulties and defects of nomadic pastoral economy rarely allow for a situation in which one nomadic group, to the cost of other groups, has a significant and constant surplus of produce. From the cultural point of view the uniform way of life of nomads, linked as it is with mobility, opportunities for migration and so on, also prevents this. From the sociopolitical point of view difficulties emerge for different reasons, varying from the weakness and undevelopment of political institutions, and the lack of a sufficiently strong machinery of coercion to the segmentary character of social organization. Peters (1968:168-9), for example, notes that in Cyrenaica strong client tribes were well able to protect themselves against noble tribes. The same is true of Mauritania (Stewart, 1972; KowalskaLewicka, 1978) and the tribes of the Western Egyptian desert. The submission of certain groups of nomads to other groups becomes somewhat more stable and long-term when interdependency on the basis of economic specialization is established between them as it was in the Sahara and, to some extent, in Arabia. However, interdependency and mutual interests themselves already limit opportunities for exploitation and the social realization of the hitler. Furthermore, strong mutual interest between 163

Preconditions

of the relations between nomads and the outside

world

dominant and subjugated groups within nomadic units or between different units only exists in cases where all sides are united by common interests with regard to the outside non-nomadic world. (In this respect the Tuareg and tribes of Mauritania are no exception.) It is in such circumstances that nomadic units become sufficiently stable to favour social differentiation, and sometimes become stratified. Thus we now come to the third and main source of social differentiation and political power in nomadic units, their relations with the sedentary agricultural and urban world. An internal development of nomadic societies does not provide sufficient opportunities for stable social differentiation. Nec vero terrae ferre omnes omnia possunt. An external factor must act as stabilizer. Such factors can be very diverse, varying from a surplus extracted from subjugated and exploited sedentary societies to social support, offered by a sedentary society (in the form of a state) to those who extend its influence amongst subjugated nomads, amongst whom the sedentary society is trying to create its social support. However, more often than not they all result in the same thing - the growth and consolidation of social differentiation in nomadic society. External links with sedentary agricultural societies, which a nomadic society strives to establish for economic reasons, at the same time are better able than anything else to encourage the growth and consolidation of social differentiation within that society; where there are the corresponding conditions they transform status differences into rank, estate and even class differences. Consequently, the people who theoretically may be referred to as a 'nomadic elite', 'nomadic nobility' or 'nomadic aristocracy' strive to establish these links not only for economic reasons, but also for social ones. Nomadic chiefdoms In the previous sections the sociopolitical organization of nomads was examined primarily from the functional-structural point of view. Now it is time for the evolutionary aspect to be taken into account. The fact that changes in nomadic societies primarily are liable to be reversible and circular can scarcely prevent us from taking the latter question seriously. If, in the end, we refute the notion that evolution is basically unilineal and unidirectional and also that social evolution can be very closely iden­ tified with biological evolution, then two facts are sufficiently clear. The first is that the unequal limits and unequal opportunities for devel­ opment which are characteristic of any specific evolutionary lines or trends (or, to put it another way, their specific evolutionary potential) do not signify complete stability or stagnation, no matter what reasons, internal or external, bring the changes themselves about. The second is that 164

Nomadic

chiefdoms

reversibility itself signifies movement from a lower to a higher level and back again; that is, it assumes certain evolutionary changes. Thus, the question of which from the point of view of evolution were the most developed nomadic societies, and what was the evolutionary level they were able to reach before they subjugated other societies or were themselves subjugated seems completely valid. From the previous statement it follows that nomadic societies can reach that stage of development which directly precedes the emergence of a state, that is, that stage which is now usually referred to as a 'stratified society' or 'chiefdom'. In order to ascertain just how applicable is the term 'chiefdom' to the nomadic societies which are being examined and to their corresponding polities, I shall try to trace the extent to which those features of chiefdoms which are most often singled out by different scholars are characteristic of them. However, this does not mean that I am always in agreement with such a characterization of chiefdoms, and the features themselves which have been singled out do not always fully coincide. 16

17

Chiefdoms

in general

1. Hereditary social differentiation, hereditary ine­ quality, presence of an aristocratic stratum (ethos) (Service, 1971:157; Service, 1975:80; Claessen and Skalnik, 1978:22). 2. Centralized government (Service, 1975:80; Claessen and Skalnik, 1978:22).

Corresponding polities +

x

3. Limited functions of supreme chief, primarily linked with legal procedure, ceremonial and external relations (R. Cohen, 1978:59).

+

4. Absence of a legitimate and coercive power to enforce decisions made by the leadership (Service, 1975:86; Claessen and Skalnik, 1978:22).

+

5. Theocratic form of government (Service, 1975:80; Claessen and Skalnik, 1978:22).

nomadic

~

"' It seems to me that neither term is completely adequate, for the first emphasizes primarily the social (or socioeconomic) aspect and the second the political (or sociopolitical) aspect. However, at this point a detailed discussion of this problem would be too lengthy and it would go well beyond the boundaries of the main theme of the book (for a more detailed discussion of the problem see Khazanov, 1979a). Apart from this, anthropological terminology is already very intricate (on this point I am in complete agreement with Pershits, 1979) and in so fur as possible I should like lo avoid a discussion of this particular problem here. " -I- signifies the presence of a corresponding feature, - signifies the absence of o n e , and x a woakei manifestation of one than is lo lie found in sedentary societies.

165

Preconditions

of the relations between nomads and the outside

Chiefdoms in general

Corresponding nomadic polities

6. Absence of effective ways of averting fission (R. Cohen, 1978:35; R. Cohen, 1978a:4; Skalnik, 1978:614-15).

+

7. 'Each nodal point in the structure has an extra replica of the officials at the centre of the system' (R. Cohen, 1978a:4; cf. R. Cohen, 1978:55).

x

8. Liable to 1975:296).

be

relatively

world

peaceful

(Service,

This table requires some commentary. 1. On the particularities of social differentiation amongst nomads see pp. 148-64. Once more I must emphasize that unequal access to key resources is not the basic reason for social differentiation amongst nomads. However, the concept of social stratification put forward by Fried which, in my view, is too narrow and rigid cannot even be applied to very many sedentary societies (see, for example, Service, 1971:157; Service, 1975:242; Service, 1978; Cohen, 1978:57; Claessen, 1978:553,588; Skalnik, 1978:604; Kha­ zanov, 1979; Khazanov, 1981). 2. As a whole political centralization is less clearly manifested in corresponding nomadic societies than it is in sedentary ones. Nevertheless, the differences are not quantitative, but qualitative, not so much structural as they are functional. Furthermore, the functions of centralized govern­ ment amongst nomads are, as a rule, narrower than they are amongst sedentaries; the main political decisions and the management of public affairs amongst the former are taken and guided by the aristocracy and only a small number of the most important decisions are taken at the highest level. It is more correct to say that there is relatively little centralization in corresponding nomadic polities and that their system of government is relatively diffuse than it is to say that they are completely lacking in centralized leadership. At the same time, in specific situations, these polities can temporarily create very centralized systems of government in order to accomplish specific aims; sometimes, as those aims are fulfilled the systems cease to exist. 'The Moor has done his duty, let him go.' Such centralization may be called dispositional. 3. In certain situations it was possible for there to be no supreme chief in nomadic polities; however, such a situation was an exception rather than the general rule. Where there was a supreme chief his functions frequently were partially similar to those of a chief in a sedentary society, for legal procedure, ceremonial and external relations. However, no less, if not more important were his other functions, for mediation in internal conflicts and military leadership. Chiefs for military purposes alone, such as the aqils amongst the Bedouin in Arabia, are an exception. But even amongst the 106

Nomadic

chiefdoms

Bedouin sheikhs have frequently acted as military leaders. Nevertheless, it is noticeable that in some nomadic societies attempts are made to split up the management of public affairs in such a way that an excessive concentration of power does not fall into the hands of just one group or person. Thus, in the Kazakh khanates of the eighteenth century the role of political leaders (khans and sultans) in legal procedure was very limited (Levshin, 1832, m: 176ff.; Chuloshnikov, 1924:208ff.). In this connection something must be said about political leadership amongst nomads. I have already pointed out that the institutionalization and centralization of power amongst nomads frequently fails to keep pace with social differentiation. This partly explains why it is that the origins of nomadic leaders can be very varied when the conditions for a dispositional centralized leadership emerge. More often than not leaders came from the aristocratic stratum of nomads themselves, frequently from a hereditary line of leaders. But this was not always the case. Mugulun, founder of the polity of the Juan-Juan in the fourth century A.D., was a runaway slave (Bichurin, 1950:184); the Shaalans came from the ranks of ordinary nomads, and the founders of the Khamseh confederation traced their ancestry back to a merchant from Shiraz (Barth, 1964:86-7). In certain areas, primarily the Muslim Near East and Africa, religious reformers and preachers, some from agricultural-urban societies, sometimes became nomadic leaders, but it seems they acted in the interests of nomads (Stenning, 1965:366-7; Gellner, 1969:4f.; cf. R. Cohen, 1978:57-8; Bonte, 1979: 179). At the same time the character of dispositional leadership is such that, in favourable circumstances and when necessity demands, it is able to emerge, even in a society in which social differentiation is undeveloped and in which an aristocracy as a specialized stratum does not exist, or else its position is very weak. The three types of authority singled out by Max Weber, traditional, bureaucratic and charismatic, seem somewhat abstract when applied to nomadic leaders. Some of these leaders, from the point of view of their origins and the character of the power they wielded, were traditional, but were able to acquire charisma through successful activity (the best example here is Jenghiz Khan). Others, on the contrary, had charisma with the help of which they could acquire leadership of the traditional forms of social organization amongst nomads. Both kinds could become bureaucratic, but only up to a point and only in a specific situation, namely after the conquest of sedentary areas and only with the help of the bureaucracy which already existed in I hose areas. But wherever this occurred nomadic polities of the chiefdom type moved over to another evolutionary level. •I. In normal circumstances absence of a legitimate and coercive power to I ()7

Preconditions

of the relations between nomads and the outside

world

enforce decisions is even more characteristic of nomadic leaderships than it is of the leaderships of sedentary chiefdoms. 5. Although very many anthropologists in the West are of the opinion that a kind of theocratic form of government is inherent in chiefdoms, I do not think there are sufficient grounds for doing so. Usually it is notions about the sacral character of the power of a chief or of his persona which are used as proof of the presence of theocracy; but clearly they are insufficient proof (for a detailed discussion of this problem see Khazanov, 1979). Moreover, even these notions apply to by no means all nomadic polities. Sacral aspects of power mostly apply to nomadic societies in East Africa. There have also been cases in the Near East and in the Sudano-Sahelian zone in which Islam has made nomadic chiefs out of individuals who have connections with religion, particularly when a need for rapid integration and centralization has emerged. Bonte (1979:180) has rightly noted that in these cases religion is an important force for yet another reason, namely its ability to overcome the limitations of tribal society. At the same time the functions of 'saintly mediators', in common with those described by Gellner (1969:78), were limited. In the Eurasian steppes, in pre-Islamic times, it was customary for Scythian kings and Turkic qaghans to perform the functions of mediation between the gods and people, but states rather than chiefdoms were characteristic of both the Scythians and ancient Turks. At any rate true theocracies, polities, that is, in which the ruling estate or class is made up of priests who take all the main political decisions, if they are to be found amongst nomads, are rare and incomplete exceptions. 6. Absence of effective ways of averting fission is even more characteristic of nomadic polities than it is of sedentary chiefdoms. 7. On account of the fact that a central leadership is more diffuse in its operation and its functions are more limited and dispositional, leadership on a lower level is not always an exact replica of the central one; however, a specific tendency towards this may be traced in a number of nomadic polities. 8. 'The relatively peaceful character' of all chiefdoms is a very imprecise feature. All that can be said of it is that corresponding nomadic polities are less 'peaceful' than some sedentary chiefdoms. It is my view that the comparison drawn above demonstrates that as a whole the term 'nomadic chiefdoms' can be used of a number of nomadic polities. However, when they are compared to sedentary chiefdoms, it becomes clear that there are some specific features about nomadic chiefdoms. First of all, their origin from the evolutionary point of view is always secondary, that is, they do not emerge sui generis, but as a result of interaction with the outside world, which can take different forms. Not only the centralized leadership in these chiefdoms, but the chiefdoms themselves to a certain extent may be called dispositional. 168

The theme and its

variations

It is for this reason that nomadic chiefdoms are usually extremely unstable, that their leadership is diffuse and decentralized and their composition fluid and impermanent. Sometimes, they structurally coincide with the higher levels of the sociopolitical organization of nomads such as the sub-tribe, tribe and tribal association, but that is not the rule. By way of example I can cite the Mongol chiefdoms of the twelfth century, or the Kazakh and Turkmenian chiefdoms of the eighteenth century and the beginning of the nineteenth, which witness the fact that nomadic chiefdoms need not necessarily fully conform to the precise levels of the sociopolitical organization in a corresponding nomadic society. In this respect I disagree with Sweet (1965:138). She characterizes relatively stable units - tribal sections (fakhd) - amongst the Bedouin of Arabia as petty chiefdoms. For Sweet, therefore, it is the structure of nomadic chiefdoms which is the basic criterion by which they are recognized. In my view, the functions carried out by these polities in the nomadic societies in question are a more reliable criterion, together, of course, with due regard for their general sociopolitical development. The theme and its variations It now remains for us to trace how the particularities of the sociopolitical organization and functioning of nomadic societies are manifested on a regional level and in historical retrospect, and also how these particularities correlate with the main types of nomadism set out in Chapter 1. However, as I have already complained, the formulation of a problem is considerably easier than the solution of it. If the example of previous ages is taken as the criterion of comparison, then nomadic societies (particularly their social organization) are best described only when relations between them and the outside world have become least typical. This fact should never be disregarded, but neither should it distract our attention from the fact that certain particularities of the social organization of nomads took shape in the modern period, under the influence of the outside world or relations between the latter and nomads, and are not traditional in the proper sense. North Eurasian type. When the social organization of such nomads as the Nentsy, Chukchi and Koriaks is examined, account must be taken of the fact that the period in which these peoples were incorporated in the Russian empire coincided with, or even preceded the establishing amongst them of large-scale reindeer-pastoralism. Amongst the Lapps reindeer-herding was also limited and regularized, by Scandinavian states, even to the extent that the latter specified quotas for the numbers of reindeer (Whitaker, 1955; Vorren and Manker, 1962). The Russian administration took measures to eliminate intergroup conflicts in this region, to gradually draw reindeer-herders into the sphere of ir>9

Preconditions

of the relations between nomads and the outside

world

monetary commodity relations and simply to protect private ownership of reindeer {Obshchestvennyi stroi u narodov Severnoi Sibiri, 1970:98-100, 385; Dikov, 1974:100-1). These measures were a contributory factor in the certain atomization of the societies under consideration and in the considerable growth of property-inequality (Kolycheva, 1956:86; Obshchestvennyi stroi u narodov Severnoi Sibiri, 1970:67, 411-13; Vasilev, 1976:327-31). Already in the eighteenth century, amongst the Nentsy, the number of reindeer in separate households ranged from 10 to 3,000 head (Zuev, 1947:32). According to data for 1926, 73.3% of nomadic Nentsy households possessed less than 100 head of reindeer, 58.5% possessed less than 50, and 40.1% less than 25 (Brodnev, 1959:77). Meanwhile by 1926-7 the ratio of the commodity output to the total output of reindeer herding in the Northern USSR had reached 18% {Obshchestvennyi stroi u narodov Severnoi Sibiri, 1970:69). However, the existence amongst reindeer-herders of redistribution and other systems of reallocation which, although they did not eliminate property-inequality, nevertheless often toned it down, should not be underestimated (Zhitkov, 1913:220; Leeds, 1965:122-3; Obshchestvennyi stroi u narodov Severnoi Sibiri, 1970:410). Although Brodnev (1959:76-7) did not know the term 'generalized reciprocity', he describes the various manifestations of it amongst the Nentsy and formulates its essence fairly exactly: 'Public opinion looked upon a refusal to help someone when they were in need of such as a very grave misdemeanour, like theft or the violation of exogamy. A person's reputation first and foremost depended on whether or not he fulfilled his obligations of mutual aid. Amongst the common laws of the Nentsy the obligation of mutual aid played an important part . . . But mutual aid is a dialectically contradictory manifestation to which there is another side. No payment could be demanded for aid given, but at the same time the person receiving aid could not refuse the same or other services to the person who was giving him aid.' Property-inequality did not occur amongst reindeer-herders in any form of firmly established social differentiation; this was both for reasons which generally are characteristic of nomads and on account of the particularities of colonial politics. Rich stockowners exercised authority, influenced choices of routes of pastoral migrations and acted as representatives of their group before the Russian administration, but this was all. A similar situation was observed amongst the Lapps (Whitaker, 1955:56; Vorren and Manker, 1962:144-5). Specific differences in social organization did exist between the different ethnic groups of reindeer-herders. The Lapps, Chukchi and Koriaks were bilateral or rather they were simply non-unilineal, while the Nentsy were patrilineal. But amongst practically all of them neither the genealogical principle of the organization of society, nor corporate descent groups, nor 170

The theme and its

variations

segmentary systems existed. Kin ties were supplemented by contractual partnership ties. One example of this kind of partnership tie is the now notorious n'ev tumgyn - 'partnership through wife' - which certain Soviet scholars have erroneously taken as confirmation of Morgan-Engel's hypothesis about primordial group marriage. Essentially, amongst the majority of reindeer-herders social organization was clearly founded on two of the basic and most stable forms of the economic plan, the separate household and the community, and also the family if the social plan proper is taken into account. More often than not a family coincided with a separate household, but this was not always the case. Communities could be made up of households, which were similar from the economic point of view, or which were different. Rich stockowners frequently resorted simply to hiring herdsmen or to letting their stock out to pasture on different conditions (Bogoras, 1904:82-3; Obshchestvennyi stroi u narodov Severnoi Sibiri, 1970:408; Vasilev, 1976:332-3; Krupnik, 1977: 80, 87). Territorial units, which included all the households and communities pasturing their stock within one area with relatively well defined geographical boundaries, were not structured in the sociopolitical sense, except with regard to general rights to grazing territory which were implemented de facto. Households wishing to move into new areas had to ask the permission of those already pasturing stock there and were sometimes refused (Obshchestvennyi stroi u narodov Severnoi Sibiri, 1970:407). Sometimes, however, nomads were aware of the existence of these units and frequently the borders of the latter coincided with the boundaries between local dialects (Dolgikh and Levin, 1951:105-6). Amongst the Nentsy (with the exception of those European groups which long ago underwent acculturation), there were exogamous clans, but the genealogical principle in their society was expressed rather weakly. They do not have clan genealogies, only legends on the origins of clans in the form of attempts to interpret links between several clans, between which marital ties were forbidden and which shared a common exogamy as if those clans had been founded by several brothers. However, no genealogical links reaching back from the Nentsy of today to supposed ancestors have been constructed. Very little research has been done into the tribal organization of the Nentsy and, indeed, into whether such organization does exist. Sometimes, on account of specific circumstances, Nentsy 'clans' joined together in larger units, but the composition of them was not fixed. Mention may also be made here of those administrative units which also were called clans. Sometimes I hey were artificially created by the Russian administration for aims which were basically fiscal. There was no linn division of pastures inside the boundaries of a given territorial unit. Pasture was looked upon as common property (Beretti, 171

Preconditions

of the relations between nomads and the outside

world

1929:21; Vasilev, 1976:335-6). Distribution of pastures was mutually arranged, without the mediation of any permanent leadership. Only towards the end of the nineteenth century and only in certain areas of the tundra (particularly the Yamal Peninsula) did separate communities acquire preferential rights to pasture. Part of the reason for this was linked to the policy of the Russian administration, for stabilization of routes of pastoral migrations enabled the administration better to control and manage reindeer-herders (Zhitkov, 1913:206-9; Brodnev, 1959:71-2; Vasilev, 1976:335; Krupnik, 1976:62-3). It is theoretically possible that without colonial power the struggle for livestock and pastures would have finally resulted in reindeer-herders creating more centralized and differentiated forms of social organization. But even so this is doubtful. The Chukchi became reindeer-herders once they had taken the greater part of the stock they needed to do so from the Koriaks and Yukagirs; however they continued to live in an unstratified society {Obshchestvennyi stroi u narodov Severnoi Sibiri, 1970:98; Vdovin, 1965:10ff.). There were simply neither the opportunities nor the stimuli for proper social differentiation amongst the reindeer-herders of Northern Eurasia. Nor, evidently, were there the stimuli for the emergence of segmentary systems, which were characteristic of nomads living in arid zones. Eurasian steppe type. By the modern period, almost all the nomadic societies and groups living in the Eurasian steppes, deserts and semi-deserts had at some time in the past experienced the extreme turbulence which the existence in a state, whether founded by their own ancestors or by their sedentary neighbours, inevitably entailed. However, amongst almost all of them periods of existence within a state alternated with periods of existence without the state and subsequent oscillations in social organization. Those scholars who regard the history of the nomads of the Eurasian steppes as a single forward development (see, for example, Lattimore, 1974; Krader, 1979) greatly simplify reality. Thus, for example, Szynkiewicz (1975:125), despite the ethnic and political changes which took place in the steppe in the Middle Ages, regards the period from the sixth century to the twelfth (and later) as one in which unilinear development, at least of some forms of social organization amongst nomads in Inner and Central Asia, took place. As a result he tends to look for the sources of the disintegration of the clan amongst the Mongols in the state of the ancient Turks in the sixth century A.D. The only trouble is that there are no facts to substantiate this view. For almost three thousand years the nomads of the Eurasian steppes confronted the outside world in the form of sedentary states, states which were often large and powerful. This is one side of the picture. The other concerns the relations amongst nomads themselves. Covering an enormous, almost continuous expanse of land in which there 172

The theme and its

variations

were no formidable barriers, nevertheless ecologically speaking the steppes did not constitute a uniform ecological zone. The western part was more fertile and richer than the eastern part. It is for these reasons that the frequent migrations of nomads towards the West were almost always accompanied by wars. Account must also be taken of the need to distribute pasture and to regulate routes of pastoral migrations, although this need was not so pressing here as it was amongst nomads in the Middle East. Nevertheless, it did exist and was frequently linked with the fulfilment of specific political functions (as distinct from the productive cycle). The result of all this was that from ancient times onwards relatively large units were very widespread amongst the nomads of the Eurasian steppes and centripetal tendencies long ago became firmly established there. It is not surprising, therefore, that as a rule the social organization of nomads in the Eurasian steppes was not limited by levels which were necessary only for its directly productive functioning. As far as can be ascertained from the sources, upper levels of social organization with certain political functions may be traced amongst the majority of nomads in the Eurasian steppes in the modern period, the Middle Ages and even in ancient times. Of course, the relative strength, stability and concrete forms of these levels were variable. According to Bacon (1958:vn, 183-5) and Krader (1963:4, 10ff.), the social organization of Turkic-speaking and Mongolian-speaking nomads constitutes a conical clan. For Bacon, who maintains that the conical clan is to be found amongst the Bedouin of Arabia and many other peoples, the principle itself of segmentation and the ambivalence of corresponding descent groups suggested by Bacon form the basis of this conclusion. Szynkiewicz (1975:114ff.) has clearly shown that this notion of their ambivalence is false. Krader goes much further. He maintains that there is genealogical ranking on the basis of primogeniture amongst the Eurasian nomads; so as to avoid any misunderstanding, I quote Krader (1963:369) on this point. 'The clan genealogy is the means whereby descent from an ancestor is demonstrated and consanguinity established, and is moreover, a means of establishing relative social rank as well. All descent lines are ranked collaterally through the birth order of the founder of the descent line, whether clan, lineage, village kin group or family. The genealogy demonstrates membership of an individual in a group where he may be personally unknown, and validates his claim to a given rank. In these societies, no one has his exact equal, everyone finds his place in a system of collaterally ranked lines of descent from a common ancestor. The ranking of descent lines among themselves and of the individual within a descent line is based on the principle of primogeniture.' 11 seems lo me that from this quotation it is clear that essentially Krader's 173

Preconditions

of the relations between nomads and the outside

world

conical clan coincides with Kirchoff s (1968) conical clan, for example as the latter is described for Polynesia. However, decidedly uncharacteristic of the nomads of the Eurasian steppes were consistent primogeniture, even more so the permanent ranking of collateral lines according to genealogical relationship to the ancestor, and the dependence of the social position of an individual upon his birthright. Here are some examples. It is well known that in the empire of Jenghiz Khan, the states of his descendants and in all other nomadic states where Jenghizids established ruling dynasties, there were no stable laws about the inheritance of power. Amongst the Kirghiz, according to Chinese sources, there was no permanent ranking of collateral lines and no inheritance of power in accordance with the laws of primogeniture (see p. 175). Amongst the Kazakhs there did exist notions about the genealogical seniority of individual hordes and their subdivisions, but these had no practical significance (Kharuzin, 1889:46; Grodekov, 1889:7; cf. Chuloshnikov, 1924:200). Amongst the Kalmucks in the eighteenth century, the aimak was an association of several khotons, headed by the zaisang which was descended from the senior line of the senior khoton. But in the next taxonomic subdivision, the anghi, the seniority of different zaisangs was not taken into account and all the anghi were considered to be equal (Nebolsin, 1852:8). More examples could be cited (see, for example, pp. 178-9 on the Pechenegs, nomads of the early Middle Ages), but in my opinion the outcome is clear. Krader's view that the conical clan exists amongst Turkic and Mongolian nomads is the result of a misunderstanding, the reasons for which still remain a mystery to me. However, it must be admitted that Krader (1963:370) does note that the Mongols of Ordos in the seventeenth, eighteenth and nineteenth centuries, and later on the Kalmucks, Kazakhs and Kirghiz began to move away from the principle of the conical clan. But the fact remains that there was no such principle. With few exceptions, various segmentary systems have been predominant amongst the nomads of the Eurasian steppes in the modern period. Characteristic of these systems was the situation in which within the same units there was an uneven number of segments in the subdivisions of one taxonomic level and even an uneven number of such subdivisions, some of which were considered closer and some more distant; therefore, a balanced opposition was impossible even in theory (Zhdanko, 1950:77-8). Furth­ ermore, all these segments differed in size. The systems themselves did not remain unchanged and altered according to external and internal circum­ stances and, moreover, in different directions. Amongst the Turkmen, in the eighteenth century and the beginning of the nineteenth, the strong tribal units which repeatedly had emerged amongst 174

The theme and its

variations

their ancestors in the Middle Ages had long ago disintegrated (Bregel, 1961; Markov, 1976:206ff.). Although there were no less than 5-6 levels in the social organization of the Turkmen, the upper levels of this organization, tribes and their large subdivisions, as a rule had no centralized political power. They represented the aggregate of small self-governing subdivisions which were linked by notions of common ancestry, but which only occasionally joined together into dispositional chiefdoms when there was a need for joint military activity. The opportunity and necessity even for nomads to move around in small groups was one reason behind the strong independence of those individual communities (oba) which were linked to the lower levels of the social organization (tire, tireh, tira, tirci). Nomadic communities independently chose the routes of their pastoral migrations. It is not surprising then that property and social differentiation were not manifested amongst the Turkmen by the emergence of a permanently separate ruling and privileged stratum. Property-differentiation in the community resulted in the emergence of rich and authoritative families, but they did not stand in opposition to the community. Muravev (1822:56) noted that in every small subdivision of the Yomut Turkmen: 'There is a chosen leader who the people obey or who, rather, the people respect for his advanced years, his deeds of brigandage or his wealth.' In some areas where the Turkmen lived large subdivisions had no permanent leaders at all, in other areas there were chiefs without any permanent real power. In the words of one observer of the Yomuts living near the Caspian Sea: 'Anyone with a little extra wealth calls himself a khan' (Galkin, 1867:31). Amongst some groups the elevation of separate lineages was observed and sometimes a title was handed on by succession; but all of this was only temporary. In the seventeenth century amongst the Kirghiz the tribes were more stable units than the tribes of the Turkmen were. Each tribe and even the large subdivisions of the tribe had hereditary leaders, biis. According to Chinese sources: 'Each bii rules his own land and has his own followers. In might and power they are equal to one another and in no way does one submit to another. When a bii dies his son and brother are set up as bii, others cannot occupy this place' (Hsi-yii wen jianlu, Buruty, p. 8, quoted in Petrov, 1961:126). Interesting changes occurred at the beginning of the nineteenth century. Individuals belonging to the Manap subdivision (named after its seven­ teenth-century founder) of the Sary Bagysh tribe gradually occupied a privileged position in their own tribe as well as in a number of other tribes (Abramzon, 1971:158). At the end of the nineteenth century the genealogies of the Kirghiz Manaps numbered 14-18 generations (Petrov, 175

Preconditions

of the relations between nomads and the outside

world

1961:122). Perhaps their segmentary system has developed in the direction of stratification. The fact that a differentiated segmentary system existed so steadfastly amongst the Kirghiz may, in my view, be explained by the way in which from the Mongol period onwards, without a break, the Kirghiz were part of different states, either nomadic or sedentary ones; moreover, at times they themselves managed to subjugate agricultural territories. At the end of the sixteenth century and in the beginning of the seventeenth, in the period when Kirghiz tribes had moved into Kashgaria and Farghana, and particularly at the beginning of the eighteenth century, their biis, with the support of their subdivisions, began to rule agricultural areas and towns. Amongst the Kirghiz the external factor which favoured social stratification was always present. The Kazakhs entered the historical arena in the fifteenth and sixteenth centuries, when their social organization was based on a stratified segmentary system. The system was the result of political developments in the preceding period when their ancestors had been incorporated into different states which emerged after the disintegration of the Mongol empire. The estate of the 'white bone' traced its descent from Jenghiz Khan and contrasted itself to all other Kazakhs, in both the social and genealogical respect, cutting across the segmentary (local) groups. However, the history of the Kazakhs from the sixteenth century to the nineteenth, with the exception of some short-term periods, is the history of the almost ceaseless decline of their political units and of a corresponding decline in the real power of the 'white bone', as represented by khans and sultans. A single khanate soon disintegrated into three Hordes (Juz), headed by elected khans who were descendants of Jenghiz Khan. The Hordes and the tribes which formed part of them were relatively stable territorial-political formations. But the power of the khans who headed them was weak, particularly as they let their control of agricultural territories and towns slip through their fingers (Barthold, 1963c:270; Pishchulina, 1969). Admittedly, stronger and more influential khans, such as Tauke, or Ablai in the thirties of the eighteenth century when the Kazakhs were at war with Jungaria, emerged from time to time amongst the Kazakhs when they were in military jeopardy. But this depended on a specific historical situation demanding a temporary unification of several large subdivisions into dispositional chiefdoms, and also partly on the individual qualities of the khans themsejves. The last such leader was Sultan Kenesary who, back in the nineteenth century, headed a very powerful uprising against the Russians and who was proclaimed Khan during the uprising. Oral tradition has it that Khan Ablai told his sons that they should never finally resolve the conflicts between the large Kazakh subdivisions, for if they were to do so they would destroy the necessary condition which was vital for preserving 176

The theme and its

variations

the khan's influence amongst them (Tolybekov, 1971:362). Khan Ablai seems to have understood the principles of segmentary organization to perfection. The most closely-knit subdivisions in the social organization of the Kazakhs were those on the lower taxonomic level, linked as they were by their joint economic utilization of grazing territory, military and other interests. The way in which the Kazakhs moved from place to place and their relations with neighbouring nomadic and sedentary states limited the initiative of the lower subdivisions of the Kazakhs more than was the case amongst their Turkmenian counterparts. The grazing territories and routes of migrations of relatively large subdivisions were specific and stable for long periods of history (Mukanov, 1974:76); when these territories and routes of migrations were broken up for external political reasons, the nomadic economy was very badly affected. Such subdivisions were headed by tarkhans, biis, batyrs, beys, aksakals and others, all of whom were members of the 'black bone'. The 'black bone' was the estate to which all the Kazakhs, with the exception of the Jenghizids and individuals tracing their ancestry back to Mohammed or to associates of the latter, belonged. The influence of such leaders, based on their authority and wealth and their links with their own subdivisions gradually increased as the influence of the 'white bone' waned (Viatkin, 1941:114). However, the top echelon of the 'black bone' did not become a closed estate and could not turn its influence and authority into hereditary power, partly because of opposition from the Russian administration. According to the 'Law on the Siberian Kirghiz' of 1822 the posts of biis were announced for election and the results of the election were approved by state power (Zimanov, 1958:196-7). The segmentary system of the Kazakhs developed from being stratified to being differentiated, and further to being non-differentiated (for the definitions of this terminology see p. 146). In my view, even if the Kazakhs had not been included in the Russian empire the social stratification in their society would still have gradually lost its significance. Only in certain cases (amongst the Mongols and Kalmucks) were the segmentary systems of nomads undermined by their prolonged existence within a state in specific conditions and, not least, by the policies of the sedentary states which had subjugated them. Mongol society in the twelfth century was stratified and had a differentiated segmentary system (see Szynkiewicz, 1976; however, he overestimates the significance of primogeniture). The policies of Jenghiz Khan and of the Yuan Dynasty and the conditions created by conquests and migrations left the segmentary system in pieces. No less a blow was dealt to it later on by the policies of the Manchu government, although lineages were H t i l l preserved or else they were reconstructed. With the creation of the empire of Jenghiz Khan, the separation of the 177

Preconditions

of the relations between nomads and the outside

world

aristocracy as a distinct privileged estate increased; and the Jenghizids created an upper, hereditary closed stratum of society amongst the Khalkha Mongols, but not amongst the Western Mongols. Thus, in its capacity as a structure-forming agent the genealogical principle of descent endured considerable changes and, most importantly, manifested itself in different ways in different estates. But amongst ordinary Mongols this principle was finally undermined only by Manchu government, aiming to create perma­ nent territorial groups in Mongolia (Grumm-Grzhimailo, 1926, in, pt. l:285ff.; Bacon, 1958:86; Bessac, 1965; Szynkiewicz, 1977). Although our information about nomads in ancient times and in the Middle Ages is fragmentary, it still confirms that the social organization of nomads then was based on one of the segmentary systems which are founded on the genealogical principle. Admittedly, in so far as most of this information relates to periods during which, or before which, nomads lived within states, we know far more about differentiated and stratified segmentary systems than we do about non-differentiated ones. It is segmentation itself which still remains somewhat of a shadow. Nevertheless, specific outlines of segmentary systems which have been steadfastly preserved in the steppe for many hundreds of years, may sometimes be observed. The Scythian royal clan had pretensions towards a divine ancestry and transformed itself into an estate, the members of which led separate subdivisions of nomads (Khazanov, 1975:191-9; Khazanov, 1978:437). In Chapter 110 of Shih Chi, Ssu-Ma Ch'ien describes the stratification of Hsiung-nu society. Ordinary Hsiung-nu 'have no name or nickname . . . [But] the duties of high officials are hereditary. Three families, the Huyan, Lan and the Xiitu who emerged later on, are regarded as the noble clans of the Hsiung-nu' (quoted in Taskin, 1968:35, 40). These clans were linked with the ruling clan of the shan-yii [kings] Liuyand'f (Hsulianti) which, essentially, remained separated into a ruling stratum. In this connection, conclusions reached by Taskin (1973:17) seem very justifiable: 'Looking at the political organization of Hsiung-nu society as a whole, it may be concluded that it is very similar to the political organization of other, later nomadic societies, particularly the Mongols.' Information about the Middle Ages is more detailed. According to data supplied by Constantine Porphyrogenitus (De administrando imperio, 37), the segmentation of Pecheneg society was founded on a system of 1:2:2:2:5. In the tenth century, the Pechenegs of the Eastern European steppes constantly raided the Russian princedom and they were drawn into Byzantine politics, but they did not constitute a single, centralized unit. 18

18

It is curious that a similar system has been recorded amongst the Karakalpaks, whose ancestors are thought by some scholars to have been the Pechenegs (Zhdanko, 1950:8211., 108-9).

178

The theme and its

variations

Each of the eight Pecheneg tribes acted independently, each was essentially a separate chiefdom, at the head of which stood hereditary chiefs. However, Constantine Porphyrogenitus (1949) alleges: 'After their death their cousins succeeded to their rule. For the law and ancient privilege have prevailed among them, depriving them of authority to transmit their rank to their sons or their brothers, it being sufficient for those in power to rule for their own life-time only, and when they die, either their cousin or sons of their cousins must be appointed, so that rank may not run exclusively in one branch of the family, but the collaterals also inherit and succeeded to the honour, no one from the stranger family intrudes and becomes a prince.' The segmentary system of the Oghuz was based on a somewhat different system. According to Rashid al-Din (1952:86-90), the upper segments of the system can be described as 1:2:3:4. According to Tolstov (1947:97), although his view remains arguable, such a system goes as far back as the time of the Hsiung-nu. Different medieval authors, Rashid al-Din, Mahmud Kashgari, Ibn al-Athir, Marvazi and Abul-Ghazi Khan, while differing on details, relate the version about Oghuz Khan and the descent of the Oghuz tribes. However, the majority of them record hereditary differentiation amongst Oghuz tribes, which they explained by differences in the seniority and purity of their descent. It would appear that not only social organization, but also the segmentary system of the Oghuz became stratified even before the Saljuq conquests. One result of those social differences was the great inequality in the ownership of property in Oghuz society, which was noted by Ibn Fadlan. If both the fact that a number of agricultural territories and towns in the Syr-Darya area became part of the Oghuz polity and that the Oghuz had links with states in Middle Asia are taken into account, none of this is surprising. In fact, a number of scholars think the polity of the Oghuz in the area around the Aral Sea and to the north of the Caspian Sea at the end of the ninth century and the beginning of the tenth already constituted a state (see, for example, Pritsak, 1953; Agadzhanov, 1969:36-42, 122ff.). I lowever, judging from the data supplied by Ibn Fadlan (quoted in Kovalevsky, 1956:127-30), central power in the Oghuz polity was weak and limited. Near Eastern type. In all respects the nomadism of the Near East is too diverse to be examined in toto. The extent and levels of social organization amongst the Bedouin of Arabia are largely determined by needs and opportunities for protecting livestock, pastures and water-sources (Rosenheld, 1965:76). The particularities of camel-herding are such that camelherders need to control large expanses of grazing territory, so the need for fairly large social units with elements of institutionalized and sometimes even centralized power arose. This power was wielded by sheikhs from individual tribes, or (heir subdivisions, and usually it was hereditary (Musil, 179

Preconditions

of the relations between nomads and the outside

world

1928:50). Differentiated segmentary systems were generally characteristic of the Arabian sub-type of nomadism, and they were most developed amongst the powerful camel-herding tribes (Sweet, 1965:13). The sheikhs kept for themselves a considerable proportion of the revenue they received from the agricultural population of the oases which they had laid under tribute, and from vassal sheep-herding tribes (Rosenfeld, 1965:78f.). They received payment for various kinds of protection they rendered in the name of the group to other groups of nomads and for safe passage through the territory of the units (Burckhardt, 1831:5, 7; Jaussen, 1908:164; Musil, 1928:59-60; Dickson, 1951:504, 562; Pershits, 1961:123; Stein, 1967:133). Sheikhs were apportioned a part of military spoils, even if they had not taken part in a raid. 'Indeed it is the rule that in the case even of raids in which the sheikh has not taken part in person, that a portion of the booty should be channelled to him' (Oppenheim, 1900, n:87; cf. Jaussen, 1908:143, 168). Although certain sheikhs reserved the best pastures for their own livestock (Philby, 1922, u, app. n; Dickson, 1951:51, 394), the greater part of their revenue and corresponding influence came from sources outside their own units, but those resources were acquired with the support of the latter (Burckhardt, 1831:95; Doughty, 1888, i:294-5; Burton, 1893, n:114; Musil, 1908:333; Pershits, 1961:139; Pershits, 1976). It is no coincidence then that within his own subdivision a sheikh's power was weak and limited, and that he had to divide part of his revenue with ordinary nomads (Montagne, 1932:71; Rosenfeld, 1965:76; Sweet, 1965:139-40). Only when grave external factors became operative did any real consolidation of political organization and centralization of power take place amongst the Bedouin. This happened, for example, amongst the Rwala at the end of the nineteenth century when the Shaalan sheikhs became very powerful after they had seized control of a considerable part of Hajj (pilgrimage to Mecca) and established complex relations with the Turkish powers in Damascus; at the same time they began to be pressured by the Rashidids' state at Hail (cf. Marx, 1977:349). However, these same external factors, particularly external danger, to a considerable extent dictated in this case the degree of social differentiation, the degree of centralization and the duration of centripetal tendencies in the society (Sweet, 1965:129f.; Rosenfeld, 1965:76). These factors were sufficient to bring about the emergence of social stratification in Bedouin societies in Arabia. However, with regard to production, distribution and, even more so, the ownership of key resources, the stratification of Bedouin society did not bring about any essential differentiation between the aristocracy and ordinary nomads. The societies of the most well-known nomads in the Sahara, the Tuareg, were more stratified, although the 180

The theme and its

variations

reasons for stratification on the whole were the same in the Sahara as they were in Arabia. However, amongst the Tuareg, social stratification developed first and foremost according to economic specialization and partly according to ethnic differences. Disregarding intermediate groups and strata, it may be presented in the following schematic form: noble tribes of camel-herders; vassal tribes, mostly herding small stock; the semi-serf agricultural population; and slaves. In my view there are two basic reasons why Tuareg society is more stratified than Arabian nomadic society. Amongst the Tuareg, camelherders, groups herding small stock and agriculturalists were integrated in one sociopolitical system (see pp. 277-8f.); this was not the case in Arabia. Account must also be taken of the marked development of slavery amongst the Tuareg, particularly the use of slaves as herdsmen (Nicolaisen, 1963:439-42), which is very uncharacteristic of nomadic societies. The exceptional aridity and isolation of the area meant that slaves could hardly escape without being caught; furthermore, the Tuareg took special measures to ensure their slaves did not escape. Thus: 'Kel Ewey and Kel Geres exchanged their captives so that the necessity of keeping a close watch on newly acquired slaves was reduced' (Lovejoy, 1976:155). Besides, in the past the Tuareg had derived considerable profit from external sources: caravan trade and the slave-trade. In no way does the social organization of the Tuareg really fit our notion of segmentary systems. In practice even noble tribes are only divided up into a few sections, and the notion of common descent is not combined with any developed genealogical principle (Nicolaisen, 1963:140ff.). No less stratified was Mauretanean society with its two privileged estates, t he military tribes of Arab origin and the tribes of Berger origin, specializing in religion and trade, with their various dependent strata and groups in which social differences partially coincided with ethnic ones. Nevertheless, subjugation of agriculturalists and the vassal-tributary dependence of one group on another were also the basis of stratification in Mauretanean society. The importance of the role played by external trade may be mentioned here as a supplementary factor (Briggs, 1960:211ff.; Stewart, 1973; Stewart, 1973a:375-85; Hames, 1979). The social organization of the majority of nomads and semi-nomads in North Africa, particularly the Arabs and those who had been arabized, was similar to the organization of the Bedouin of Arabia. Differentiated segmentary systems predominated and the society had become stratified. Hereditary leadership of a tribe was held by a certain lineage in one subdivision; there were, however, no strict rules about succession. The sheikhs were the military leaders, they dealt with external relations, directed legal proceedings, distributed grazing and watering places, and IKI

Preconditions

of the relations between nomads and the outside

world

regulated pastoral migrations (Marcais, 1913:254-7; Brunschvig, 1947:100ff.; N.A. Ivanov, 1963). Nevertheless, amongst the nomads of North Africa, the main factors in social stratification were not relations between different groups and units of nomads and certainly not relationships within these groups and units, rather they were the relations between nomads, on the one hand, and the sedentary population and state, on the other. It is no coincidence that, as a general rule, the makhzan (government-supporting, tax-extracting) tribes were the most stratified. The social organization of Berber nomads, primarily the Zanata, was, according to Ibn Khaldun (Ibn Haldun, 1925, m:179ff.; cf. Golvine, 1957:136-7), similar to that described above. But already amongst Berber semi-nomads and, even more so, amongst those engaged only in herdsman-husbandry there were a number of peculiarities in the social organization, the most particular being that from a social point of view there was less differentiation (Golvine, 1957:23ff.). Sometimes the segmentary systems of the Bedouin of Cyrenaica and of the Berber semi-nomads of the High Atlas, or of pastoralists there engaged in distant-pastures husbandry, are called classic systems of segmentary lineages. The latter are well described by Gellner. 'Berbers are not stratified into aristocrats and commoners, clans do not possess permanently dominant sheikh families' (Gellner, 1969:64). However, I would not wish to exaggerate their egalitarianism, at any rate not from the social point of view. Although the word tribe is one which should be used somewhat conditionally with regard to the Berbers of the Central Atlas, nevertheless amongst them there did exist a need for upper levels of social organization (Gellner, 1969:49, 90-1). Limited needs for integration and centralization did not favour social differentiation within the society. But still these needs did exist and they gave rise to a remedy which corresponded adequately to their aims. Functions of hereditary leadership are in many nomadic societies concentrated in specific lineages. Amongst the Berber of the High Atlas these functions to a considerable extent were concentrated in lineages of 'saintly' mediators, who were outside the genealogical framework of the society. These 'saints' partially fulfilled those functions which in most nomadic societies are left to the aristocracy (Gellner, 1969:78). It is also clear that the limited social differentiation amongst the Berbers of the Atlas fully ties up with the absence of those external factors which in other nomadic societies have encouraged differentiation. To a great extent this also applies to the Bedouin of Cyrenaica, an area 19

19

In the Near East the separation of power between tribal leaders and religious lineages is not limited to the Central High Atlas. A n analogous picture is to be observed in Cyrenaica, the. Hadramaut and the Northern Mountains of the Y e m e n (Gellner, personal communication).

182

The theme and its

variations

where there were practically no peasants and towns played no significant role (Evans-Pritchard, 1949:44-5). Admittedly, there were 'free', and client, dependent tribes (Peters, 1968:175) and, moreover, the degree of dependence of these tribes varied; however, it was insufficient for any marked social differentiation to emerge, at least within the 'free' tribes. However, the various subdivisions of tribes were still headed by sheikhs and the functions of leadership were fixed by succession in one and the same lineages (Evans-Pritchard, 1949:59-60). Such a social organization may be called segmentary (see pp. 144-8), but scarcely egalitarian. The social organization of the Somali, the main representatives of the Northeast African sub-type of the Near Eastern type of nomadism, has been characterized by Lewis (1955; 1961) as a segmentary lineage system. However, in several respects, some of them mentioned by Lewis (1961:297-302; 1965:344), it differs from the classic examples. Some of the particularities of Somali social organization are the unusually great significance and depth of genealogies, the lack of balanced opposition between segments of one taxonomic level - they are differentiated according to strength, numbers and influence, the presence of the tribe (clan, in Lewis' terminology) which is the bearer of corporative ownership of key resources with clearly marked genealogical boundaries and hereditary power, albeit weak, appointed to the sultans and fixed in certain lineages. Apart from this, amongst the Somali there are dependent groups which do not have full rights and which are not part of the genealogical framework of the social structure. Thus, potential preconditions for the development of stratification in Somali society did exist, but they did not develop. Why was this? Evidently, because there was neither the call, nor the stimuli for them to do so. The ecological situation of the Somali was such that it was really almost impossible for them to organize a stable distribution of pasture and to regulate routes of pastoral migrations, which is one way nomads can consolidate political power. Neither did the practice of raiding agriculturalists become very wide­ spread amongst the Somali. Far more common were raids by certain groups of nomads on other groups. This was linked to the shortage of pasture and to the insignificant role played by agriculture. In addition, agriculturalists had religious protection and some of them were part of the genealogical structure of Somali society. Nevertheless, in the past the power of the sultans was somewhat greater. It is indicative that sultanates tended to emerge not within nomadic society, but on its borders where sultans could find supplementary sources of revenue and power, for example, control of trade routes. Nomadic tribes acknowledged their power when an external factor, war with Ethiopia, came into force (Lewis, 1961:209). IK3

Preconditions

of the relations between nomads and the outside

world

Thus, historical circumstances did not encourage the Somali to join together for aggressive expansion or for defence, nor did they stimulate internal social differentiation. The Somali advance to the south was one of gradual penetration, rather than simultaneous expansion. The Somali played a very active role in the attempts of the Adal Sultanate in the fifteenth and sixteenth centuries to conquer Ethiopia, but these attempts were unsuccessful. After the sixteenth century there was little hope for a successful struggle against Ethiopia which was stronger. Nevertheless, it is indicative that there is more social differentiation amongst the southern Somali, who are seriously engaged in agriculture. Middle Eastern type. I made the point in Chapter 1 that, in contrast to the nomads of the Eurasian steppes, the nomads of the Middle East were in considerably closer, more permanent and more day to day contact with the sedentary population and states of the region, and that the interdependence between nomads and the sedentary world was greater here. From the geographical point of view, for the nomads of the Middle East the 'outside world' was not outside in the real sense of the word. On the one hand, their pastures and pastoral routes were usually within the territory of one state or another, if only purely nominally so; on the other hand, no complete spatial differentiation could be made between nomadic pastures and pastoral routes and sown fields. Gellner (1973:2) has called this symbiotic nomadism, consisting of a partnership of three - nomads, agriculturalists and townsmen. Routes of pastoral migrations and distribution of pastures had to be very exactly and strictly coordinated. De Planhol (1969:88ff.) suggests that the latter characteristic acquired particular significance in Iran in the late Middle Ages; this was when nomads had occupied all the ecological zones suited to them and began to need more pastures. As a result, large nomadic units emerged, which to some extent still exist in the present day. In certain areas of the region, primarily in Iran, the aforementioned characteristics were also very closely interconnected. At the same time interrelations between various groups of nomads played a lesser role in the Middle East than they did in the Eurasian steppes. Barth (1962:348) has singled out three forms of sociopolitical organiza­ tion amongst the nomads of the Middle East in the recent past, between which, as is usually the case, there are also transitjonal forms. The three forms are: (a) 'atomistic, kin-based camps, without political ties with other nomad camps'; ( b ) ' a group of camps tied together by patrilineal descent in a segmentary lineage system, without chiefs'; (c) 'a group of camps, segments of which may approximate lineages in structure, but defined as a unit through common recognition of a supreme central chief. Evidently, the first, atomistic form, which may be observed, for example, amongst the Turkish Yoruk (Eremeev, 1969; Bates, 1971:123f.; Bates, 1974) is the result of events in the present day or the recent past, the fruit of 184

The theme and its

variations

changing relations between a sedentary population with its state machinery and weakened nomadic groups. In brief, this is a result of the disintegration of the latter. In the Middle Ages such isolated communities of nomads would have been either destroyed by agriculturalists or by other nomads, or else they would have been incorporated into their sociopolitical organiza­ tion. The second non-differentiated form was not widespread amongst the nomads of the Middle East. It would appear that in part this was for the same reasons, although the policies of the state could lead to opposite results, an increase in social differentiation (see p. 213ff.). Various forms of differentiated and stratified segmentary systems were most widespread amongst nomads in the Middle East at least until the recent past, and in part are still to be found in the present day. There are a great many such systems. At one extreme are such nomadic units as the Boyr Ahmad and, particularly, the Behmai and the Taiyebi, amongst whom there are hereditary leaders and chiefs; however they are relatively decentralized and isolated from the outside world (Fazel, 1973:138). At the other extreme are such powerful and centralized units as the Qashghai, the Basseri and certain Afghan and Baluch groups, which are very closely linked with the outside world and which are stratified into various privileged, ordinary and unprivileged strata. An aristocratic stratum sometimes ceases to be a part of the genealogical cliche of the units which it heads, and follows distinct estate rules of kinship and descent (see, for example, Narody Perednei Azii, 1957:280; M.S. Ivanov, 1961:88-92; Barth, 1964:71ff., 130ff.; Trubetskoi, 1966:128, 139). The segmentation of the various systems amongst nomads in the Middle East appears to be very limited. Essentially, only two features, the genealogical principle and the fact that they are multi-level are widespread. But, as is the case in the Eurasian steppes, structural equivalency is, as a rule, absent; thus, the segments of the same level differ in size and in the number of segments of a lower level which they contain. Neither is the structure-forming role of the genealogical principle absolute. It does not embrace all the subdivisions, from top to bottom, of the given society, and various subdivisions can be regarded as genealogically unrelated (see, for example, Fazel, 1973:134 on the Boyr Ahmad; Barth, 1964:49f. on the Basseri). It should be pointed out that in the Middle East, as indeed is the case everywhere, the greater the general social stratification and centralization of a nomadic society, the more limited and specialized is the genealogical principle in its capacity as an organizational force. Admittedly, the ruling stratum, at least at the beginning, is recruited by genealogical proximity to the leader or central core-group. But first, individuals surrounding leaders, bodyguards, for example, may have different genealogies. Secondly, the IK5

Preconditions

of the relations between nomads and the outside

world

important social and property positions of the aristocracy act as a magnetic force for the establishing of contractual relations of dependence (for example, vassal or patron-client dependence) both with individuals, and with entire groups, which are not connected by common descent (see, for example, Barth, 1964b: 17-19; Spooner, 1975:172-3; Salzman, 1978a:130ff. on the Baluch). External factors play an extremely important role in social differentiation amongst the nomads of the Middle East, as they have done amongst the nomads of the Eurasian steppes. The implications of this statement are, in effect, very broad, for in the Middle East even the regulating of routes of pastoral migrations and the preservation of grazing territories are, in many ways, connected with the relations between nomads and the sedentary population. In this process the immediate relations between nomads and the sedentary population and state are the basic and decisive factors. It is interesting that the economic and ecological differences between the two types of nomadism are reflected in different manifestations of the same factor. Amongst Eurasian nomads the character of the relations between nomads and the outside world were the determining factor; amongst the nomads of the Middle East it was their position within sedentary society itself. However, of course these differences, like all the others, should not be overestimated, for local and historical variations should also be taken into account. Afghan tribal units included both kin and non-kin tribes, the latter usually being vassals of the former (see pp. 162-3); actually, in terms of the notion of the genealogical kinship of all Afghans, these vassals were minor kin. Furthermore, in Afghan tribes there were various categories of dependent populations, basically consisting of agriculturalists and collec­ tively referred to by the Persian term 'hamsaya' (literally 'under one shadow', 'common shadow', i.e. neighbour, etc.). It was considered that they did not belong to the tribal organization (Elphistone, 1819, i:274ff.; Ferdinand, 1969:144). All of this enabled social stratification to develop amongst the Afghans. By the Middle Ages many tribes had already singled out a subdivision (khan khel) in which the hereditary positions of chiefs were fixed. Such, for example, were the Saddozais amongst the Abdali (Durrani). They were endogamous, enjoyed a number of privileges, personal immunity and the rules of blood feuds did not apply to them (Elphistone, 1819, n:116). Ahmad-Shah, founder of the Durrani state was himself a Saddozai (Reisner, 1954:49f.). At the same time the social organization of many other Afghan tribes is characterized as acephalous, without fixed leadership (Ferdinand, 1969:137). East African type. The ecological conditions for the functioning of nomads in East Africa are such that neither the need, nor the opportunities 186

The theme and its

variations

for distributing grazing-territories and regulating the routes of pastoral migrations exist (Forde, 1970:21-3; Monod, 1975:128). At the same time the factors which favour political integration and centralization are not very clearly manifested. It is thanks to the aggregate of these factors and a number of other ones that even community links have developed relatively weakly amongst nomads in East Africa. Quite apart from institutionalized and centralized power, multi-level social organization and such characteris­ tics as the genealogical principle, hierarchy of descent groups and segmentary systems have not developed here to the extent that they have in the Near East. Not only amongst the Samburu, but also amongst the Rendille is membership of a clan linked with the notion of common descent (Spencer, 1965:xxiv-xxv, 50, 76-7; Spencer, 1973:78). The Jie do not retain the memory of any ancestors before their grandparents (Gulliver, 1955:76). The system of social ties amongst East African nomads is very flexible. It is not only the relatives of an individual father, but also those of his mother and those acquired through marriage which play an important role. Partner-relations of the contractual type, bond-friends, stock-associates and stock-friends, have acquired great significance (Gulliver, 1955:197f.; Spencer, 1965:27-8; Dahl and Hjort, 1976:263). All of these provide the individual and his household with sufficient means for satisfying both their economic and social needs. The latter involves, first and foremost, the creation of mechanisms needed for the redistribution of livestock, without which no proper atomistic production can function normally. A system of values (accumulation of stock for acquiring wives, sacrifice, etc.) also operates in the same direction, encouraging reciprocity. The system of age-grades and age-classes amongst East African nomads evidently had a polyfunctional significance, quite apart from the causes for, conditions and place in which it emerged. It is not surprising that in the past the defensive and offensive potential of a society were concentrated in that system (Gulliver, 1955:12; Spencer, 1965:96, 100; Goldschmidt, 1965:4034; Alpers and Ehret, 1975:491). Furthermore, age-classes permitted the age- and sex-division of labour, marriage relations and the domestic cycle to be regulated in such a way that the individual household was guaranteed the stock and work force it required in the most efficient manner (Spencer, 1965:299; Dahl and Hjort, 1976:75, 256). All this made it possible for the system to overcome atomistic production. Thanks to the various redistributive mechanisms that have emerged in response to socioeconomic needs, the property-differences amongst East African nomads, which are sometimes considerable (see, for example, Gulliver, 1955:99-110 on the Turkana and the Jie; Spencer, 1965:2 on the Samburu; Klima, 1970:15 on the Barabaig; Spencer, 1973:40 on the 187

Preconditions

of the relations between nomads and the outside

world

Rendille), did not result in any stable social differentiation. Of course, there were differences in prestige and status, there were East African counter­ parts to Melanesian big-men (Dahl, 1979:278), just as there were elements of the institution which in anthropology is designated by the rather vague term gerontocracy, but on the whole none of these became hereditary. The expansion of the Turkana in the second half of the nineteenth century did not cause any essential changes in their social organization (Gulliver, 1955:7). Admittedly, amongst the Barabaig (Klima, 1970:83, 95, 98) there were heads of clans and even a supreme chief whose positions were genealogically secured in specific lineages according to the laws of primogeniture. However, at least in the present time their functions are primarily ritualistic. In the past these chiefs were also military leaders, and their power was somewhat greater. In this sort of ecological, economic and social situation the subjugation of one set of pastoral societies by another was practically impossible. There was only the struggle for stock and pastures, when the latter were in insufficient supply, in the form of individual raids. Opportunities for turning stock into immovable property (land) or another form of investment (Dahl, 1979:279) were non-existent. Spencer (1965:xvii-xviii) insists that the difference between the Samburu and the Maasai consists in the fact that the former were surrounded by the same sort of pastoralists as themselves, while the latter, the Maasai, lived alongside agriculturalists on whom they could carry out lucrative raids. It was amongst the Maasai that social differentiation acquired an hereditary form with the singling out of hereditary priests, laibon, who were descended from one clan and who also fulfilled secular functions as leaders of society. However, we have very little precise information about the pre-colonial history of the Maasai. It would appear that when various groups of the Maasai began to fight amongst themselves in the nineteenth century, the Maasai became weaker (Alpers and Ehret, 1975:492). Opinions differ about the nature of their society in this period (see, for example, Jacobs, 1975), thus it is too risky for us to come to any real conclusions from such material. Amongst the Nandi there was a supreme priest, the orkoiyot, who in part fulfilled the functions carried out by chiefs in other societies. But Huntingford (1953:50) explains this by an increasing need for centralized leadership, linked to the growth of the importance of agriculture in Nandi society. Relations with the outside world were as a whole limited amongst the East African pastoralists that are being examined in this book. Non-pastoral products were obtained either by way of exchange, or if the pastoralists themselves were engaged in some agriculture. Oscillations between nomadic and semi-nomadic pastoralism were, evidently, a permanent feature of the way of life here (see, for example, Galaty, 1981). It is hardly 188

The theme and its

variations

surprising, therefore, that no deep and stable social differentiation developed amongst them. In this respect they are very different from the earlier pastoral migrants in the Interlacustrine (see Chapter 5). The societies of East African pastoralists and nomads are sometimes called egalitarian without any precise specification of what is meant by this broad and indistinct term. Sometimes the earliest in the evolutionary scale and/or most primitive societies are called egalitarian. In this case, if the pre-history of East African nomads is taken into account, their 'egalitarianism', even from the purely theoretical point of view can only be called secondary, because nomadism itself is a relatively late branch of food-producing economy. But even their secondary egalitarianism gives rise to specific doubts, although not on account of the reversibility of social processes amongst nomads, to which reference frequently has been made. In the end, a specific level of development is inherent in any society. By what path and from what preceding level this is attained are another matter. The main doubts centre on whether those definitions which, with certain variations, are usually used of egalitarian societies apply to East African pastoralists. According, for example, to Fried (1967:x, 27f.) one of the most important characteristics of such societies is generalized reciprocity, which does not demand the return of what has been given. However, amongst nomads and pastoralists of East Africa, it is not generalized, but balanced and partly even non-balanced reciprocity which operate as redistributive mechanisms; they were to some extent familiar with social differentiation and inequality, but in very few instances have these become hereditary. The attribute 'non-stratified' (cf. Lefebure, 1979:2) seems to be a more suitable one for describing them than egalitarian. However, one drawback is that this term's meaning is only negative. There is as little homogeneity about the nomads of the Sudano-Sahelian zone from the social point of view as there is from the economic one. As a whole their social organization is more similar to that of the nomads of the Near East than to those of East Africa, but social differentiation is less developed amongst them than it is amongst the nomads of the Near East. Amongst the Humr Baggara a fairly typical system of segmentary lineages is to be found, with only an informal leadership. Even in the pre-colonial period, when raids to procure slaves were a lucrative occupation, there was no hereditary leadership (Cunnison, 1966:8-10, 42, 189-90). A mongst the Wodaabe Fulani of Bornu, the Fulani of the Benue Valley of Nigeria and some other groups only small agnatic descent groups (lineages) with informal or weak leaders, not clans consisting of several such lineages, were of any significance until the echoes and after-effects of the Uthman dan l o d i o movement reached them (Stenning, 1965:375-85, 393-4; R. Cohen, 1978a:l50 4). 189

Preconditions

of the relations between nomads and the outside

world

Amongst the Kababish, one lineage turned into an almost endogamous stratum with the monopoly of political power and rights to preferential use and control of natural resources. But this kind of stratification did not arise out of economic or social conditions, rather it was directly linked with specific historical events in the colonial period and with the policies of colonial administration (Asad, 1970:102, 158f.). We do know, it is true, that in the Sudan throughout most of the Middle Ages and right up to the nineteenth century, nomads frequently created various states or they were involved in the creation of such. Only in these situations did their social organization undergo any more or less rapid and significant transformation. However, all of these states were basically agricultural (Stenning, 1959:13ff.; Stenning, 1965:366-7; Smith, 1971:1659; Horton, 1971:110-11; R. Cohen, 1978a:144ff.). My final conclusions coincide with those of Burnham (1979:353). In nomadic societies in Africa south of the Sahara, the growth of social differentiation is linked either to their conquest of sedentary societies or to the sedentarization of nomads themselves. Internal development amongst nomads on its own is insufficient for growth of social differentiation. High Inner Asian type. Social organization in the nomadism of this type is not homogeneous because the Kirghiz of the Pamirs do not differ in any essential way from their kinsmen in the region where there is the Eurasian steppe type of nomadism. Relatively little is known about the sociopolitical organization of the nomads of Tibet. Furthermore, the fact that over many centuries nomads here were politically subordinate to a sedentary state (states, more precisely) could not fail to affect their social organization. In certain periods their dependence was greater than it was in others and in the central areas it was more acutely felt than it was in the Northeast, on the borders of the Eurasian steppes. Kozlov (1947:215) came across a nomadic unit of the Goloks in Kam (Qam) at the end of the nineteenth century, in which there was an hereditary aristocracy and which recognized neither Tibetan, nor Chinese authority. But, as a whole, lack of political independence was the most important characteristic of the nomads of Tibet for many centuries (Lattimore, 1967:212-13). The question of ownership of key resources amongst the nomads of Tibet is not very clear. It has been said that in certain places they had to rent pastures (Kychanov and Savitsky, 1975:126; Reshetov, 1975:210). At the same time Ekvall (1954:46) has noted that pastures here are usually looked upon as belonging to the tribes and are shared out amongst the communities which make up these tribes. Possibly the situation varied according to place and time. The genealogical plan of the social organization of the nomads of Tibet cannot be easily traced - at least I have not been able to find any detailed 190

The theme and its

variations

information about this in the literature. Evidently it was less important and less developed here than it has been amongst the nomads of the Eurasian steppes or the Near East. In common with all types of nomadism, in the economic plan the social organization of Tibetan nomads is based on individual households and communities (Ekvall, 1968:25ff.). The sociopolitical plan proper most clearly consists of the family in its different variants and the primary kin group. But the clan is described by Ekvall (1968:28-9) as '. . . a somewhat amorphous lineage grouping . . . presumably related to a common ancestor from whom it takes its name . . . somewhat like a clan in the process of decay.' Members of one clan could be scattered in different tribes. Clearly tribes were primarily political organizations. But whereas in Central Tibet they were greatly modified by their frequent inclusion as distinct districts in the general administrative system of a state, still at the beginning of the twentieth century in the Northeast tribes enjoyed considerable autonomy, reluctantly submitting either to the Tibetan or Chinese governments (Ekvall, 1968:29). In these areas sometimes even tribal units, resembling dispositional chiefdoms, were formed. There are indications that in several tribes the power of chiefs became hereditary (Reshetov and Iakovlev, 1975:210), but it would appear that more often than not this power was not very great, even in outlying areas. The rigid estate-class division of the sedentary population of Tibet was not at any rate characteristic of the nomadic population. Southern Asia. In Chapter 11 expressed my doubts as to whether Indian pastoralism can be singled out as an independent type of nomadism. These doubts are substantiated when the social organization of pastoralists in Southern Asia, or rather the lack of anything so specific in comparison with the surrounding agricultural population, is examined. With rare exceptions, pastoralists do not comply with what in India is customarily called 'a tribal society'. All those particularities which might have been part of their social organization in the past have disappeared. On the contrary, from the ethnic and social points of view pastoralists are now identical to agriculturalists; in common with the latter, pastoralists are included in the caste system (and in Ka jasthan in such high castes as the Raikas, Jats and Rajputs) and also in the system of village communities, and there are no lands or pastures which they own either separately or differently from agriculturalists. This short survey of the social-political relations which exist (or have existed) in nomadic societies in the different types of nomadism does not in my view contradict the preliminary conclusions reached above. Despite the, opinion of certain scholars (see, for example, Sahlins, 1968:37-9; Johnson, 1969_159-60; Markov, 1976:278-313), nomadism I'M

Preconditions

of the relations between nomads and the outside

world

should not be regarded as an autarkic economic system, even less should it be regarded as a distinct, closed sociopolitical system corresponding to a specific stage (or stages) of evolution, a system with its own internal laws of social functioning and development. As is made clear by the history of the origins and spread of nomadism as a whole, nomadism is not a specific stage in universal evolution in the way it has been defined by Carneiro (1973:97-100). Furthermore, there are no characteristics of sociopolitical organization which are to be found amongst all nomads, or which are to be found exclusively amongst nomads (Spooner, 1973:3; Tapper, 1979:45). Finally, we can specify no one evolutionary level which can accommodate all nomads. The most we can do in this respect is to take the lower and upper levels of their sociopolitical evolution and functioning as politically independent societies, and then only with a number of essential reservations. Equally unpromising, in my view, is the attempt to explain all variations in the sociopolitical organization of nomads, as well as the general level of development of separate nomadic societies directly, or exclusively, by ecology (see, for example, Eberhardt, 1952:69-72; Rubel, 1969; Salzman, 1971; Swidler, 1972; Spooner, 1973; Dahl, 1979; cf. Pastner, 1971a; Khazanov, 1973; Frantz, 1978; Lefebure, 1979a; Bourgeot, 1979), although, of course, the ecological factor is one of the most important determinants. We can take any ecological factor which is characteristic of different nomadic societies and easily show how these societies do nevertheless differ from one another in the sociopolitical respect. For example, relatively stable routes of pastoral migrations are characteristic of the atomistic reindeer-herders of North Eurasia, also of the considerably more socially differentiated nomads of the Middle East, while irregular routes are characteristic of various nomads in the Near East and pastoralists in East Africa. The species-composition of the herds of the Somali is similar to that of the Bedouin of Arabia, and the Tuareg's similar to that of the Teda. Separate pasturing of individual species is character­ istic of the Somali, the Rendille, the Tuareg and the Bedouin. Even if the sum total of all the various ecological factors is taken into account, it is still not sufficient to explain the social, and even more so, the political differences between different nomadic societies. This is because the sociopolitical specificity of these societies is by no means only a question of passive adaptation to their habitat. Far be it from me to belittle the work of the scholars mentioned above and of those other scholars who are primarily concerned with ecological factors. It is thanks to their research that our understanding of nomadism has moved so much ahead. But the historical process does not yield to simplification. In the past the influence of ecology on this process was often not given sufficient consideration. However, as a basic method of research into social 192

The theme and its

variations

and political systems and their changes, ecological determinism is just as unpromising as economic, cultural- or historical-materialistic or, indeed, any other determinism. It is then for this very reason that I am equally unconvinced by the attempts of Marxists and Marxist-oriented scholars which, citing Marx and Godelier, analyse nomadism and its evolution by proceeding in the main from its supposed internal, independent socioeconomic development. For example, Helfgott (1977:39; cf. Bonte, 1974) tries to construct a special nomadic socioeconomic formation. Lefebure (1979:11), on behalf of a group of French anthropologists, is more cautious. Of the nomadic societies he considers to be class societies he poses the following question: 'Can our characterization of the socioeconomic specificity of nomadic pastoral societies lead to a definition of a single mode of production or are we obliged to recognize several of them, or ought we to speak in terms of an economico-social formation?' However, Marxists have themselves always insisted, and with rare exceptions still insist, that in every formation there is a distinct mode of production which is formed out of the aggregate of productive forces and social relations of production. Terminology new to anthropology, such as superstructure and infrastructure, which Godelier (1978) tries to introduce, does not essentially change anything. The Marxist view is that every mode of production has a qualitative distinctiveness in comparison with others. In so far as one of the basic postulates of Marxism is that the mode of production contains specific socioeconomic relations, those Marxists, before describ­ ing the nomadic mode or modes of production, should demonstrate in what way they are different from all other modes of production. But so far no one has done this at all convincingly. In my view this is because extensive pastoralism, which forms the economic basis of nomadism, is not a mode of production, but only a distinct form of economic activity. Nor is nomadism itself an autonomous system in any one of its basic parameters. In this respect pastoral nomadism stands in one typological line alongside agriculture, hunting and fishing, but certainly not alongside the so-called Asiatic mode of production, slave-owning, feudalism or capitalism, if we are going to follow Marxism and regard these as distinct stages (formations) in the universal-historical process or, contrary to that point of view, are going to regard them as the result of local historical development, concrete manifestations of multilineal evolution. I should like to remind those Marxists who adhere to another view of pastoral nomadic societies of an aphorism of one famous Marxist, Engels (1969:56): 'If I include a shoe-brush in the unity mammals, this does not help it to get mammary glands.' 20

•'" I (should be pointed out that Godelier himself says it is wrong 'to confuse forms of the division ol labour with modes of production' (Godelier, 1978:765), as do (or did) certain other French Marxists.

193

Preconditions

of the relations between nomads and the outside

world

Refuting the ecological approach, members of the 'Ecologie et anthropo­ logic des societes pastorales' group in Paris suggest that nomadic societies, owing to their internal development, are basically capable of independently evolving from primitive, 'egalitarian' to class societies (see, for example, Bonte, 1974; Bonte, 1975; Bonte, 1975a; Bonte, 1977; Lefebure, 1979; Hames, 1979). And in all fairness it must be pointed out that these scholars acknowledge that such a point of view can be discussed further. 'Criticism is invited', one of them wrote not long ago (Lefebure, 1979:13). I think that the point of view set out in the present chapter shows quite clearly that my view is really rather different from that of my French colleagues. Nevertheless, I accept their kind invitation and should like to make some additional points here. As their point of departure on the path of historical evolution they take unstratified 'egalitarian' nomadic societies, conceived of in the model of East African pastoralists of the Jie or Karimojong type. Owing to mechanisms of redistribution property-differences do not remain constant in these societies. Bonte (1974:76-7; 1975:388) and after him Lefebure (1979:4) see many similarities between them and the ancient Germans. Characteristics of both '. . . seem to us to constitute an original historical evolutionary path of which, following Germanic societies between Caesar and the great invasions, still undifferentiated nomadic pastoral societies may be regarded as a typical product' (Lefebure, 1979:6). It seems to me that this opinion is both dogmatic and incorrect. The 'original' nomadic societies referred to are conceived of in the same way as Marx thought of the ancient Germans in his Grundrisse - production was carried out by individual households, but ownership of key resources belonged to the society as a whole. And on this basis the society itself is called a community, 'community founded - prior to any form of productive activity' (Lefebure, 1979:4-5). I shall not here touch on the problem of what Marx himself had in mind. This would involve an unnecessary textual analysis of the sources of Marx's terminology, which anyway is far from being consistent, of the problems of translation, and a special digression into what exactly Marx meant in 1857 by the community (evidently not always the same as Bonte means by it). But Bonte and Lefebure ignore the fact that corporative ownership of key 21

22

2 1

2 2

The factual material on which such notions are based remains unclear. A n y anthropologist taking upon himself the task, apart from Marx's Grundrisse, to acquaint himself at least superficially with contemporary archaeological and historical literature about the ancient Germans, or who at worst simply reads Tacitus with care, will have no trouble in discovering that their society in the first to the third centuries A . D . cannot be called anything but stratified. B e t w e e n 1857-8 and 1979 German studies have moved forward, so that notions about social organization, land tenure, the economy and the general level of development in ancient German societies have changed a great deal. Marx differentiates Gemeinde, the community proper, and Gemeindwesen, a very broad and unclear notion which he uses in different senses.

194

The theme and its

variations

resources, quite apart (let us say) from the ancient Germans and 'egalitarian' nomads, is characteristic of many other nomadic societies, ones which from the social point of view are considerably more differentiated. In that case, if we are consistent, we must call them communities as well. We can, of course, call them such, but in that case the evolution from 'egalitarian' nomadic societies to non-egalitarian ones clearly has not taken place in the direction conceived of by certain members of H'Equipe.' At the same time they have failed to notice real producing communities, including those amongst East African nomads. Development of productive forces in a pastoral economy, which entails the establishment of permanent property-inequality, is thought by members of 'I'Equipe' to be a prime moving force in the development of social differentiation. '. . . The formation of non-egalitarian relations of produc­ tion in nomadic pastoral societies seems to us generally to occur in periods of rising labour-productivity and accumulation within domestic units, thereby leading to a sharpening of internal contradictions in the society in question. The fundamental contradiction within the specific structure of egalitarian relations of production, namely the contradiction between the domestic framework of productive work and the community conditions of reproduc­ tion, thus seems to be determined by a contradiction between relations of production and forces of production' (Lefebure, 1979:8). Further, these supposed contradictions result in the formation of groups and strata which exclusively own the means of production (livestock, pastures), and groups and strata without them. Economic differences develop into social differences, and these social differences develop into class differences. In all fairness, it must also be pointed out that those scholars do acknowledge the influence of the outside world on nomads. In a number of specific case-studies they have made many valuable and interesting observations about this subject which, moreover, sometimes contradict their own basic views. Nevertheless, on the theoretical level these observations are clearly given insufficient weight. As a result facts come into conflict with theory. The works of the Paris Marxists do not make clear how the supposed increase of productive forces in a nomadic pastoral economy is achieved. Evidently, it is no coincidence that they are unable to cite one example of internal development in a nomadic society from an egalitarian to a class society. Instead Bonte (1975) compares the non-stratified pastoralists of East Africa with, what is in his opinion, the 'transitional' stratified society of the 23

" For example, B o n t e (1975a) rightly emphasizes that the development of the agricultural sector in the e c o n o m y of the Tuareg Kel Geres was an important factor in their evolution into a class society. All that remains unclear is why he then considers that the formation of a class o( dependent and exploited agriculturalists of non-Tuareg origin was the result of the internal development of Tuareg society (cf. Burnham, 1979:353-4).

l

I>5

Preconditions

of the relations between nomads and the outside

world

Maasai and, further, the class societies of Interlacustrine region. He suggests they all represent different stages in the internal evolution of East African pastoralists and, moreover, religion does not legitimize the emergence of classes, but plays a role in their actual formation. (Although Bonte does not refer to Max Weber, in the last respect the parallels between his views and those of the author of the 'Protestant Ethic' seem to be very marked.) But Bonte does not testify to the main part of his hypothesis: whether productivity of labour in the pastoral economy in Rwanda and Ankole is really higher than it is amongst the Maasai, and whether amongst the Maasai it is higher than amongst the Karimojong and Turkana. This is hardly accidental since the concrete facts of the history of East Africa stand in striking contrast to his hypothesis. Let us leave the Maasai aside. The question of the level of social differentiation amongst them in the pre-colonial period is far from clear, furthermore it is doubtful whether social differentiation was exclusively linked to internal factors (Jacobs, 1975). However, no matter how contradictory the opinions of different scholars about the emergence of the Interlacustrine states, one thing remains clear: it was connected with the interaction between pastoral and agricultural groups, integrated in one way or another in one society (see pp. 290ff.). Consequently, there is clear evidence here of the external factor. The same problem to some extent has long been a subject of discussion amongst Soviet anthropologists. Some suggest that there are no fun­ damental differences between nomadic and non-nomadic societies, that both develop according to the Marxists' 'laws' of the universal-historical process, which are the same for mankind as a whole, and, moreover, these anthropologists always give preference to internal development (see, for example, Zlatkin, 1971; Lashuk, 1973; Fedorov-Davydov, 1973; Pershits, 1976). Others, on the contrary, believe that the evolution of nomads is inseparable from their relations with the outside world and in the end is subject to the specificity of the latter's evolution (Khazanov, 1973; Khazanov, 1975; Markov, 1976). It is noticeable that between these two above-cited approaches to the problem, which purely conditionally can be called the Marxist and the ecological, there is, despite all the differences, one common feature. Ecologically oriented anthropologists explain the social particularities and sociopolitical development of nomads as internal ecological adaptation. The 24

2 4

Markov's stand on this question is an extreme o n e . H e believes that the level of development amongst nomads in Asia practically always was identical. In his description of this level lit: employs a notion introduced by Morgan and Engels, 'military democracy'. Even conquest of sedentary states very little, and only temporarily, in his view, altered the pre-clnss, socioeconomic relations which existed amongst nomads (Markov, 1976:279, 307-H, 312).

1%

The theme and its

variations

French scholars mentioned above and some of their Soviet colleagues explain them as evolutionary changes on account of internal contradictions. In both these approaches there are aspects worthy of serious attention. To some extent I have already touched on these in the present chapter. Nevertheless, my final conclusion is very different from the conclusions reached by the other scholars to which I have referred. In my opinion, from the sociopolitical point of view, nomadism is not merely the economic adaptation of pastoral society to its habitat. At the same time, neither is it a distinct socioeconomic formation, or a distinct mode of production. The level of sociopolitical development in nomadic societies and corresponding changes in this level are very largely determined by the specificity of the societies' relations with the outside world, and with the individual particularities of the latter.

197

4 Modes of nomadic adaptation to the outside world

Two aspects of the adaptation of nomads to the outside world, the economic and the sociopolitical, are far more easily differentiated in a theoretical analysis than they are in practice. Nevertheless, facts remain facts. All types of nomadism are non-autarkic and cannot, therefore, function in isolation. Differences exist in the ways and means of adaptation, but not in needs and aims. As a general rule, the more differentiated a given nomadic society from the social point of view, the more actively it strives for contacts and interaction with the outside world. Admittedly, in the majority of cases the feedback claim is also true, for intensive and multilateral contacts with the outside world enable social differentiation and/or property inequality to increase amongst nomads. The effect of the feedback link is far more fitting an explanation for the given phenomenon than is a one directional cause-effect chain. All that has been said also applies to many semi-nomads. Because their economy is more complex, semi-nomads feel these same needs less acutely than nomads do, but they still do feel them. Thus, the choice of specific ways in which a nomadic society can adapt to the outside world depends on the economic needs of all its members and on the sociopolitical aspirations of individual strata and groups. And, of course, when we are looking at the ways in which nomads adapt to the outside world we must always take into account the specific opportunities and limitations which nomads have to contend with. The outside world does not usually act as a passive background where nomads are concerned, rather it acts as an active force in this interaction. The combination of all of these circumstances can lead to the most diverse results, so that even the same ways in which nomads adapt to the outside world are manifested differently. Nevertheless, it seems to me that an attempt to single out the basic ways is worthwhile. Sedentarization The first and most radical way is sedentarization, an increase in agricultural production and the development of handicrafts. But here nomads face many 198

Sedentarization difficulties and hindrances, both psychological and social. Sedentarization means that stereotypes of thinking, behaviour, a traditional system of values and a traditional way of life must be broken. The notion that the sedentary way of life is the antithesis of nomadic life is not only a reflection of the differences between the conditions in which the two exist, but probably also an integral part of the ideology of nomads, sometimes consciously or unconsciously cultivated by a nomadic aristocracy. As one manifestation of the universal opposition between 'us and them', on the functional level this ideology at once plays an ethnically and culturally integrative role in nomadic society proper and a differentiating one with regard to the sedentary world. Furthermore, it is fundamental in the negative attitude towards the sedentary world. Once such an attitude, which by nature is secondary, has taken root, it stands as an obstacle in the path of sedentarization. However, this obstacle still is not always taken into consideration by administrators in power, anxious to rid themselves as quickly as possible of the problem of nomads in the contemporary world. Social obstacles to sedentarization, although on a different level, are in many respects connected with psychological ones. The sedentarization of an individual family sometimes tears the nomad from a traditional system of social ties and deprives him of important lines of defence (see pp. 82-4). 'Often (as in Turkey) nomads who settle must interject themselves into ethnically different and initially hostile communities' (Bates, 1971:114; cf. Bates, 1974; Monod, 1975:141-2; Dahl and Hjort, 1979:29). Mass sedentarization is frequently linked to the specific disintegration of a nomadic society and an essential transformation in its social organization. Not all strata and groups in a given society derive equal benefit (or disadvantage) from the transformation, and for this reason not every one is interested in it to an equal extent. It is noticeable that private-ownership tendencies with regard to land increase amongst nomads as they become sedentary (see, for example, Markov, 1976:224 on the Turkmen; Erdniev, 1970:38-40 on the Kalmucks; cf. Spooner, 1973:15). Of course, both the psychological and social obstacles confronting sedentarization are surmountable. There are many historical examples which substantiate this fact quite unequivocally. When the founder of the Liao Dynasty in the tenth century marched onto the territory of Mongolia and visited the ruins of the former Uighur capital on the Orkhon, he suggested to the Uighur ruler of the town of Ganzhou that he return to the former lands of the Uighurs. However, although only eighty years had passed since the Uighurs were forced out into East Turkestan, they had already grown accustomed to an agricultural-urban life and had no wish to return to the nomadic life of their forefathers (Barthold, 1963e:498-9). The greatest obstacles which nomads faced and those which they found most difficult to overcome were ecological. In so far as pastoral nomadism 199

Modes of nomadic

adaptation

to the outside

world

primarily developed as the most effective economic alternative to agriculture in ecological zones unsuited to the latter, so in a zone of pastoral nomadism agriculture was only feasible in territorially limited enclaves and in marginal areas. However, even in marginal areas agriculture was frequently less profitable than it was in a traditional agricultural zone, well suited to cultivation. Furthermore, it is in these areas that the borders between nomadic and agricultural societies are at their most changeable, depending on climatic changes, the general historical situation in the region as a whole, and the specific alignment of forces and corresponding policies of the more powerful side. (On the changing situation in this respect over the centuries in the South of East Europe see Khazanov, 1973; in Middle Asia sec Barthold, 1963b: 120; on the borders between China and the steppe sec Lattimore, 1967; in North Africa see Julien, 1956; Julien, 1956a; cf., however, Asad, 1973:66-7; in Arabia and adjacent countries see Pershits, 1961:39ff.; in Turkey see Bates, 1971:121; in the Sahel see Gallais, 1977; sec also pp. 78-80 above.) All of this meant that in some cases nomads were forced to transfer to agriculture because their pastures were reduced and/or because of corresponding policies of sedentary states, and in other cases agriculturalists had to give their land up to pasture or themselves turn to nomadism. In most of the territory occupied by nomads agriculture is economically able to compete effectively with pastoral nomadism only if supported by expensive irrigation-works and a large amount of capital investment, and the introduction of specific farming cultures; and even then it is far from everywhere that this is possible. With rare exceptions, up to the present day irrigation and capital investment have been almost out of the question. And even in the twentieth century the eagerness of certain politicians frequently forestalls the creation of corresponding stimuli and opportunities for sedentarization. Successful, complete and mass sedentarization was only possible through migrations into other ecological zones favouring agriculture. Such migra¬ tions are well known in history. In some cases they were mass, simultaneous migrations, in others gradual infiltration. Sometimes they were the result of conscious and purposeful action, but considerably more qften they were (he natural consequence of the interrelations amongst nomads themselves and with sedentary societies in a specific historical situation. There is no one single reason why nomads migrate. Consequently, there can be no one explanation for their migrations. Contrary to the opinion of Toynbee (1935:15), the migrations of nomads are caused not only by climatic changes or the weakening of sedentary civilizations, but also by a number of othei factors, including those rooted in the particularities of the functioning,", of nomadic societies themselves at the given historical moment. In such cases 200

Sedentarization sedentarization has frequently been readaptation to a new ecological and social environment. After the Mongol invasion of Middle Asia a number of Mongol and Turkic nomadic groups moved there (Petrushevsky, 1977:132). At first this led to a decline in agriculture and an increase in pastoral nomadism in the area. However, gradually the majority of the new arrivals in this ecological zone which was new to them began to transfer from pure nomadism to semi-sedentary pastoralism and even to sedentarization. Similar processes frequently occurred in Middle Asia in the pre- and post-Mongol periods (see pp. 78-9, 249, 261). The last to follow this path were the nomadic Uzbeks after Shaibani Khan had taken Middle Asia from the Timurids (P.P. Ivanov, 1958:72). A similar phenomenon in the history of the Near East is well known. Over the course of many hundreds of years those nomads who were forced, for different reasons, to leave Arabia frequently became sedentary once they found themselves in a new ecological zone. Aside from those in biblical times, some of the most ancient examples of such sedentarization are the Tanukh of Yemeni origin who, in the third century A . D . , migrated into the fertile area to the west of the Euphrates and there created the Princedom of the Lakhmids (Hitti, 1956:81). Cases are also known of Sahara Berber tribes (including such large ones as the Zanata) and Arab tribes becoming sedentary. Frequently it was those nomads who had suffered defeat and been forced to give up their pastures to their victors who became sedentary (Capot-Rey, 1953:181, 282-4; cf. Coon, 1931:408 on the Rif tribes; Fvans-Pritchard, 1949:45-6 on Cyrenaica). The Middle East is no exception. Mention has already been made of the sedentarization of Arabian Bedouins on the territory of Iran (see p. 104). For many centuries pastoralists and nomads from the mountainous areas of contemporary Pakistan and other areas have migrated into the Indus valley and settled on lhe fertile lands here, and sometimes have even penetrated southern areas of the subcontinent (Doni, 1974:111-14; cf. Leshnik, 1972:158ff.). In the lifteenth, sixteenth and seventeenth centuries, and later, many Afghan tribes began to go over to agriculture after they had seized agricultural lands, or else migrated into free lands suitable for cultivation. Ways in which nomads migrated were different, but the results were similar (Reisner, 1954:46, 256). Frequently transition to agriculture was necessary. Thus, on the plain by the lower reaches of the Kabul-Darya and in the mountainous valleys of Swat and the Panjikora, seized by Khakhay tribes at the beginning of the sixteenth century, there were not the wide pastures necessary for largescale nomadism. So pastoral nomadism had to be supplemented with agri­ culture (Reisner, 1954:46-7). The Baluch who migrated into Baluchistan and displaced the local 201

Modes of nomadic

adaptation

to the outside

world

population there themselves began to go over to agriculture; in Makran this was already happening in the thirteenth and fourteenth centuries. This process was particularly intense in eastern Baluchistan where the Baluch seized agricultural lands which had long been cultivated by the local population (the Jats, Rajputs and others); the latter were exterminated, driven out, or became slaves or dependent tenants (Pikulin, 1959:26; cf. Salzman, 1978a: 132). Yet another example in another region is the Somali. When nomads from the north of Somalia penetrated the more fertile areas in the south, their economy became more complex, combining agriculture with pastoralism (Lewis, 1960:227). However, relatively complete and mass sedentarization, by whatever paths this is accomplished, overcomes the one-sidedness of nomadism by the very fact that it renounces it. Thus, once again we return to the problem of the adaptation to the outside world of those nomads who, however, remain nomads, and the ways in which their adaptation occurs. Trade and mediatory trade Differentiation must be made between the two different types of trade between nomads and the outside world: (1) direct exchange and trade, basically with agricultural and urban societies; (2) mediation or participa­ tion in the trade between different sedentary societies, and the services and other contacts linked to these. Direct exchange and trade between nomads, on the one hand, and between agriculturalists and townsmen, on the other, essentially involve a number of obstacles which are hard to overcome, particularly if the long history of these relations is taken into account. An economy founded on pastoral nomadism is not infrequently an economy of relatively high involvement in the market. Among arat families (arats = immediate producers) of pre-revolutionary Mongolia the proportion of produce exchanged amounted to 30 per cent of total output (Markovska, 1973:288). As a general rule, because their economy was specialized and one-sided nomads were more interested in trade than sedentary societies in which the economy, with individual exceptions, was more complex, and, consequent­ ly, more autarkic. Natural economies were characteristic of sedentary societies in ancient times and in the Middle Ages. Furthermore, trade with the restless world of nomads always was linked with possible involvement in military confrontations and pillage on the part of nomads. It was harder for nomads to sell a surplus product (when they had one) within their own society with its undiversified economy. It must also be remembered that in their trade with sedentary societies nomads were as interested in handicraft production as they were in agricultural products. Despite their repeated 202

Trade and mediatory

trade

efforts and the diversity of methods adopted to do this, nomads never managed to organize handicraft production in a proper quantity and of a proper quality in their own society. At best they managed palliatives, but again this was with direct or indirect assistance from the outside world. Ibn Khaldun well understood all of this, as is made clear in his pronouncement cited in the previous chapter. This does not mean, of course, that agriculturalists were completely uninterested in trade with nomads or that they could not derive any benefit from such trade. Facts indicate that the contrary is true. It was just that, for nomads, trade with agriculturalists was a necessary and vital concern, particularly when they were unable to secure for themselves supplies of agricultural and handicraft products by other methods. Agriculturalists could, in principle, get by without trading with nomads (cf., for example, Bates, 1971:115 on the situation in Turkey). Things were somewhat different with towns as centres in which were concentrated handicraft-industries and trade. However, in the majority of cases towns were an integral part of more stable systems within the sedentary world, and they were primarily committed to the maintenance of these systems. Trade with nomads could sometimes bring in greater profits, but as a permanent factor it was too unstable to monopolize the trading and sales of towndwellers in any long-term perspective. Of course, there have been exceptions to this rule, especially in the Near East. In Morocco in the Middle Ages: 'Towns were isolated from the countryside except when they served as market-centres for the bedowins. Then the main tie between town and country was an uncertain symbiosis between pastoralists and mer­ chants, rather than the more stable domination of landlords over peasants' (Lapidus, 1975:35). Therefore, the first obstacle consists of the economic differences between nomadic and sedentary agricultural-urban societies. The second is once again linked to the specificity of nomadic economy. From studying the ancient and medieval historical sources, a scholar may derive a false impression about the economic potential and stability of nomadic economy. In these sources there is a quite considerable amount of information about the interest of nomads in trade with the sedentary world and about the great quantity of livestock and pastoral products they supplied to the markets of the sedentary world. Inevitably the question then arises as to how a nomadic economy could produce the necessary surpluses for trade when instability is one of its permanent characteristics. The correct answer to this question requires that three different sets of circumstances be taken into account: the nature of historical sources, the specificity of nomadic economy, and the property and social differences in nomadic societies. The information about trade between nomads and the sedentary world in (he available sources is, as a rule, fragmentary, incomplete and sometimes 203

Modes of nomadic

adaptation

to the outside

world

only relates to those periods and years even when there was an increase in trading. Frequently, it was because of such increases that trade caught the attention of an ancient or medieval author. Because a nomadic economy functions cyclically, it is very likely that in many such cases the sources reflect a favourable cycle of growth which temporarily created the surpluses necessary for trade. Furthermore, property and social differences in nomadic societies often lead to the situation in which, although all members of the societies want to be involved in exchange and trade with the sedentary world, opportunities for becoming involved are different. Sometimes trade is implemented to the detriment of the redistributive mechanisms in nomadic society. On the other hand, for the nomadic aristocracy and also for rich stockowners, trade can serve as a canal through which to stabilize and strengthen their economic and social positions. It is for this reason that amongst traded items sometimes luxury goods appear, which Lattimore (1967:xlii) is inclined to regard as all but the basic items for trade between nomads and the sedentary world. There are social motives behind the desire of the nomadic aristocracy to make contacts with the outside world, but distinct economic motives, which often are not in the interest of nomadic society as a whole, are also involved. Indeed, these social and economic motives are usually so closely intertwined in practice that it is hard to separate them. Finally, it is certainly not always the case that in their trade with the sedentary world nomads sell only the surplus of the goods they have produced. Sometimes they have to forgo a part of a product they themselves need. An exchange of livestock and pastoral products for agricultural products can worsen their dietary system (see, for example, p. 81) and even the long-term perspectives for reproduction in the herd, but nevertheless be necessary as long as it allows them to hold out on less calorific products in inauspicious seasons and cycles (cf. Dahl and Hjort, 1976:179-81). Indeed, there are known cases of nomads keeping so few cattle that they have been forced to sell their children and wives (see, for example, al-Umari, al-Maqrizi, and al-Ayni on the Tatars of the Golden Horde — quoted in Tizenghausen, 1884:231, 235, 436, 513; data about the Kalmucks in the eighteenth century - Palmov, 1927:125; on the Kazakhs at the beginning of the nineteenth century-Viatkin, 1941:216). Salzman (1978a: 131) notes that in bad years the Baluch simply had nothing to trade with. In the nineteenth century a series of direct and indirect measures devised by the Russian administration, interested in maximum supplies of Kazakh livestock for Russian markets, and the usurious work of merchants coming into the steppes led to the large scale removal of livestock from there; this stock was not replaced by a natural increase of numbers in herds and nomads were ruined (Apollova, 1976:346ff.; Markov, 1976:201). 204

Trade and mediatory

trade

Thus, on the one hand, not only the specialization, but also the instability of pastoral nomadism made nomads more interested in trade with the sedentary world; but, on the other hand, this same instability did not favour production of regular surpluses which could have been sold regularly at the market and satisfied the needs of the entire society. Trade with the sedentary world is an indispensable attribute of any type of nomadism. However, the significance of this trade is not everywhere the same, and specific regional and even local characteristics are very marked. I have already cited data which shows that, with regard to exchange and trade, reindeer-herdsmen of North Eurasia are no real exception to other nomads (see pp. 72-3, 81). The main difference is only that for them trade was essentially the only way of acquiring necessary agricultural and handicraft products. They had no other alternatives. Thus reindeerherdsmen very much wanted to trade, even under conditions which were very unprofitable for themselves. A special trait of the Chukchi was that exchange with the Eskimos was of very considerable significance for them. The Eskimos were hunters of sea mammals, and this balanced the dietary system of the Chukchi. In addition, the Eskimos in part fulfilled for the Chukchi the same role as the handicraftsmen did in sedentary agriculturalurban societies further south (Krupnik, 1976:47). In the Eurasian steppes the distinctive character of trade between nomads and the sedentary world was determined primarily by two sets of circumstances, one of which was ecological, the other political. The fact that agriculturalists and nomads utilized separate ecological zones created specific spatial difficulties for trading. Sometimes livestock and other goods had to be driven and carried over great distances. It is noticeable that it was nomads who played the active role. Barthold (1963b: 123) has even written that '. . . everywhere there has been trade between nomads and sedentary peoples - on the borders of China, the Muslim world and Russia - it is the nomads who have driven their herds to the centres of frontier trade without waiting for traders to come into the steppes'. In fact, this is not exactly the case. It is well known, for example, that Soghdian merchants played an active role in the Turkic Qaghanat (Empire). Ibn Fadlan has testified that Muslim merchants penetrated deep into lands where the Oghuz lived (Kovalevsky, 1956:126-7). Nevertheless, more often than not the situation was as it is described by Barthold. Individual figures look very impressive. Thus, in the autumn of 1527, 20,000 horses were driven into Russia, 'from the Nogai', to be sold; in 1529/30 it was 80,000; in 1530/31 it was 30,000; and in 1532/33 it was 50,000 (Zimin, 1972:221). The Kazakhs of the Little Horde sold to Russia between 1798 and 1802: 10,919 horses and 649,282 rams; between 1803 and 1807: 4,314 horses and 406,715 rams; between 1808 and 1812: 282 horses and 205

Modes of nomadic

adaptation

to the outside

world

318,208 rams (Viatkin, 1941:215). However, for the reasons mentioned above, I would not accept these figures as absolute and, most importantly, that they remained as such over long periods of time. The considerable year-to-year fluctuation which can be observed in the figures cited here is in itself sufficient evidence. It is not impossible that some migrations and conquests made by nomads were initiated by the simple desire of the latter to be a little closer to their much needed markets. Such was the policy of the Scythian king Atheas back in antiquity (Khazanov, 1975:242; Khazanov, 1978:430). Also worth citing here is the view of Barthold (1963:467-8) that the Mongol conquest of Middle Asia was accelerated by the shortsighted actions of the KhwarazmShah, Muhammed, who closed the trade routes from Maveraunnahr into the steppes. Possibly, the desire to be somewhat closer to Chinese markets can explain the shift of the centre of political power amongst the Mongols from the shores of the Tola, Orkhon and Kerulen into the Chahar areas in the south in the period between the death of Esen and Dayan Khan's consolidation of power (Zlatkin, 1964:62). It is curious that available sources almost always show nomads as the side most needing trade and sedentaries as the side benefiting most from this trade. Maqdisi made a special note of the low prices for meat on the northern borders of Maveraunnahr. According to al-Istakhri, in the tenth century Khwarazm was bound by its wealth to trade exclusively with nomads, and the trade between the two was concentrated in the northern part of the country with its centre in the town of Gurganj (today Kunia-Urgench). Thanks to this trade the emirs of Gurganj became so powerful that at the end of the tenth century they seized the southern part of the country and transferred the title of Khwarazm-Shah to their own dynasty (Barthold, 1963b: 124). Later on the Saljuq sultan, Sanjar, recorded especially in an edict: 'Benefit is derived from their [i.e. nomads] wares and goods, these are a reason for the increase in the prosperity, contentment and profits of settled peoples' (quoted in Materialy po istorii turkmen i Turkmenii, 1939:314). However, the sedentary states bordering onto the nomads of the Eurasian steppes usually regarded trade with nomads as an instrument of external politics, a way of applying economic pressure. In China trade with nomads was usually directly carried out or regulated by the state. During the course of this trade an identical situation recurred with remarkable regularity. China strove to stop or limit trade with nomads when the latter were loth to admit that they had been subjected, and nomads acquired the right to trade by using arms. In the peace-treaties which China had to conclude with the Hsiung-mi, the latter insisted that China pledged to open up markets at frontier posts. In the words of Ssu-Ma Ch'ien (Shih Chi, Ch. 110, quoted in Taskin, 1968:51), 206

Trade and mediatory

trade

'. . . notable for their greed, the Hsiung-nu valued the markets at the centres of frontier traffic and value Chinese articles; but China was also interested that there should be markets at frontier posts in order to satisfy their desires'. In 1431 the Emperor Hsuang-zong spoke openly about his reasons for allowing trade with nomads. 'China started to trade not because she has no horses or cattle [incidentally, there was a marked lack of horses there A.K.]. On the contrary, the barbarians are in every way dependent on China. Therefore the court's toleration of trade is at the same time a display of conciliating humanity' (quoted in Martynov, 1970:233). In this case the experience of many centuries was rarely of any useful ser­ vice to the Chinese. This was not because, as the famous aphorism says, the only lesson to be learned from history is that there is no lesson to be learned from history. It was just that the economy, sociopolitical organization and ideology of China were so different from their nomadic counterparts that it was simpler for China to risk war than to shake her own foundations. But even in those states where attitudes towards trade with nomads were more 'liberal', governments never had political considerations far from their minds. By a law of 1212 the Tangut state Hsi Hsia (Ta Hsia) forbade individuals who went to the Tatars to make dishes and other utensils for them. The punishment for breaking this law was three years of hard labour (Kychanov, 1977:58n.3). Shaibani-Khan forbade merchants from Samarqand, Bukhara and Khwarazm to sell wares to the Kazakhs (Akhmedov, 1965:110). 'Several times there was issued the most august edict that the population of Turkestan should have no trading deals with Kazakh merchants and that no reciprocal visits and journeys of merchants between them and the inhabitants of these lands should take place. Thus several times in certain areas of Turkestan and in the towns of Khwarazm it was decreed that Kazakh merchants should be robbed. His majesty the Khan had all manner of wise considerations in mind when he issued this edict' (Fadlullah ibn Ruzbikhan Isfahani, 1976:101). For nomads in the Near East, particularly in the Sahara and in Arabia, trade was of even greater significance than it was for nomads in the Eurasian steppes and also took place under more profitable conditions. Both the fact that in many areas of the Near East sedentary life was concentrated around small oases and also that nomads have been militarily and politically strong in this region for most of the last fifteen hundred years has meant that, to a considerable extent, the maintaining of trading contacts has depended on nomads (cf. Coon, 1976:63). Furthermore, the narrowness of the internal market and the vicissitudes of external trade meant that urban handicraft industries and trade were partially oriented towards satisfying the requirements of nomads (de Planhol, 1966:273; see also Evans-Pritchard, 1949:44 5 on Cyrennicu; Capot-Rcy, 1953:217-19, 227 on the Sahara; 207

Modes of nomadic

adaptation

to the outside

world

Pershits, 1961:51 on Arabia). According to the calculations of Smilianskaia (1979:94) in the seventeenth and eighteenth centuries about 5% of the urban population of Syria was involved in supplying goods to nomads. Nomads acquired millet, dates and other agricultural products, also handicraft goods, from oases without any problem. The inhabitants of the oases always needed meat, butter, cheese and other products of pastoral economy (see, for example, Nicolaisen, 1963:63, 218 on the Sahara; Pershits, 1961:55 on Arabia). Furthermore, towns in the Sahara and in Arabia were as trading-centres heavily dependent on trade with nomads. In Arabia nomads had their permanent trading-centres to which they gravitated every summer (Wallin, 1850:36, 46; Doughty, 1888, i:123; Philby, 1922, i:170; Dickson, 1951:49; Pershits, 1961:55-6). In addition, itinerant traders who sell goods directly to the Bedouin in the camps of the latter have survived up to the present day (Awad, 1962:334; Stein, 1967:85; cf. Burckhardt, 1831:154f.). However, it was not only in the towns and oases of Arabia that there was a great demand for the products of pastoral economy, particularly wool and camels (Marx, 1977:347). But it was professional traders rather than nomads who derived most profit from the export of these products (Pershits, 1961:42-3, 56ff.). Even so for the same age-old reasons supplies were limited, although up until World War One the Rwala alone annually sold up to 30,000-35,000 camels (Oppenheim, 1939, i:102). Consequently, the caravan trade was extremely important for many nomads in the Near East. Nevertheless, whenever possible sedentary states in the Near East never missed an opportunity to use the interest of nomads in trade as a means of political pressure. Blunt (1879, i:69) testifies that the Bedouin depended on the good will of the pasha when they arrived in towns in the spring to sell their livestock and to buy agricultural products and clothes, and that they had to buy the right to trade. In the Middle East nomads and sedentaries were particularly closely linked to one another, because of both history and ecology. It is not surprising, therefore, that close economic interdependence was established between them back in ancient times and that regular exchange and trade became a permanent factor in their relations. For example, in Afghanistan the basis of these relations was not only the interest of nomads in grain, but also agriculturalists' need for. milk products. The situation which existed until recently has been described in a fair amount of detail by many scholars (see, for example, Schurmann, 1962:255, 261-2; Barth, 1962:346ff.; Barth, 1964a:72; Barth, 1976:74; Ferdinand, 1969:134ff.; Spooner, 1975:180f.; Glatzer, 1981). However, I should not like to say without reservation that the situation as described by these scholars is also typical of the past, when the economic side of relations between nomads 208

Trade and mediatory

trade

and sedentary populations has been surpassed and mediated by the military-political side. The nomads of the Horn of Africa are involved in regular exchange and also in barter and credit trade with agriculturalists (Lewis, 1955:71, 78). Judging from certain data (see, for example, Gulliver, 1955:6-7, 211; Van Zwanenberg and King, 1975:149; Dahl, 1979:266), for East African nomads and semi-nomads exchange with the sedentary population is more important than the way it is presented in many analyses. In the eighteenth century the Maasai, despite their raids on the Kikuyu, at the same time traded with them and aquired ironmongery from them (Alpers and Ehret, 1975:493-4). Amongst the nomads of Tibet non-economic ways of overcoming economic specialization were limited. Thus different forms of trade became very widespread. These ranged from barter to money deals, from the exchange of pastoral products for grain and handicraft goods at local markets or directly with trade partners from amongst agriculturalists to the delivery of thousands of sheep, cattle and horses to the markets of West China (Kiuner, 1908:65f.; Kozlov, 1947:372; Downs and Ekvall, 1965:175, 180; Ekvall, 1968:69). Long-distance caravan-trading and the supplying of caravan traders were sometimes more profitable for nomads than was the direct exchange of pastoral products for agricultural products and handicraft goods. Since trade involved goods which were in the main produced in sedentary countries, the pastoral economy was spared any tension involving necessary exchange or sale of its products. The role of nomads in this trade was extremely diverse ranging from mediating trade to the transportation of loads, the sale or renting out of transport animals, the conducting or safeguarding of caravans, or simply to payment for unimpeded passage, etc. In this case it was not only their geographical location, mobility and ownership of transport animals which was to the advantage of nomads, but also their psychological attitude towards travelling, migration and move­ ment beyond the boundaries of territories they traditionally occupied. It was not by chance that all the great overland trade routes of antiquity and the Middle Ages were pioneered by nomads or with their participation. Account should also be taken of one other factor. In contrast to many attitudes in ancient and medieval societies both in the West and the East, amongst nomads trade and connected professions stood high up on the scale of values and enjoyed great prestige. Caravan trade was of very considerable importance for nomads in the Near East; There is data which shows that semi-nomads participated in caravan trade as far back as the second millennium B.C., that they used donkeys as transport animals and acted as guides and escorts (Kupper, 1957:15, 119). It also suggests that the 'Spice Route', which connects the 209

Modes of nomadic

adaptation

to the outside

world

Mediterranean with Southern Arabia, and through there with India and East Africa, was established by nomadic camel-herders in the first millennium B.C. (Rathjens, 1952-3:283). Mediation in the caravan trade was of considerable significance to the ancient Bedouins of Arabia. Its disintegration was a severe blow to their well-being and was one reason for the important changes which took place in the history of the peninsula. It is not impossible that the camel-breeding specialization of the nomadism of the Near East and even the spread of the camel to other types of nomadism were partly connected with the requirements of the caravan trade and the benefits which nomads derived from this trade (Bulliet, 1975). Amongst the Bedouin of Arabia in the modern period caravan transporta­ tion and carrying of hajj have served as a supplementary source of income (Awad, 1962:334). However, townsmen specializing in this business have been able to compete successfully with the Bedouin (Pershits, 1961:33-4, 56, 57-8, 60). The caravan trade was very important to the nomads of the Sahara, amongst certain groups of which it became one of the most important sources of revenue, sometimes even the main means of existence (Capot-Rey, 1953:217-18; Bovill, 1958:53; Nicolaisen, 1963:209-12, 216, 218-19; Stewart, 1973; Hames, 1979:377ff.). It is quite possible that it was the struggle for control of caravan routes which was sometimes the real cause behind the events leading to migrations, conquests and even the emergence of nomadic states (see, for example, Levtzion, 1971:127-8 on the Almoravids). In the Middle East its significance was more limited and primarily involved certain specialized groups of Afghan nomads, powinda (powindeh). Already in the thirteenth and fourteenth centuries they supplied transit trade on the routes linking India with Persia and Middle Asia. They were able to do this because caravan routes coincided with the routes of their pastoral migrations so that the powinda were separated from their stock and families only when the latter were at their summer pastures. They rented out camels, were employed as stock drivers, received payment for guarding caravans and themselves engaged in mediatory trade. A group of Lohanis, which wintered in the Indus valley and had summer pastures on the Ghazni plateau and on the borders of Khurasan gradually turned the mediatory trade between India and Iran, and then that between India and Middle Asia, into a secondary (and in certain cases even the basic) occupation of their own. In the sixteenth century the powinda twice a year made the journey from Bukhara through Qandahar and the Gumal pass to India. Even in the nineteenth century, when this sort of trade was declining, a considerable part of the Dotanis, the Mian Khel and other tribes were powinda. However, although nomads in Central Afghanistan are engaged 210

Trade and mediatory

trade

(or were engaged) in mediatory trade to the present day, the subjugation of Hazarajat at the end of the nineteenth century brought about in Afghanistan what Ferdinand has called 'the commercialization of former pure pastoralists'. Some groups of nomads have turned into itinerant merchants par excellence (Reisner, 1954:50, 65, 68; Thesiger, 1955:319; Schurmann, 1962:255, 262; Ferdinand, 1962:125; Ferdinand, 1969:138ff.; Azimdzhanova, 1977:44). Other groups of nomads received revenue from the caravan trade, mainly for the protection and safe passage of the caravans through their territory (see, for example, Coon, 1976:195 on the Baluch). The significance of the caravan trade was more limited in the Eurasian steppes. At best it was here a supplementary source of revenue for some groups of nomads. Admittedly, it had existed from ancient times. The Scythians brought Greek wares to the foothills of the Urals (Herodotus, IV.24), and Strabo noted of the Sarmatian Aorsians that they grew rich by transporting goods on camels and trading with Indian and Babylonian wares (Strabo, XI.V.8). The Great Silk Route from China to the Middle East, the Black Sea and the Mediterranean was the most well known. But although nomads sometimes profited from this route, more often than not this was when they themselves had founded powerful and extensive states and were able to control the territories through which it passed. Otherwise, the Chinese, for example, could alter their route in certain territories and thereby bypass nomadic areas. Thus the Silk Route ran south of the Tien Shan, from Khami through Kashgaria and further on to Farghana and Samarqand, that is through the oases of East Turkestan and the agricultural areas of Middle Asia. This was despite the fact that the route through the nomadic territories to the north of the Tien Shan turned out cheaper, as camels could be pastured along the way there in rich pastures (Lattimore, 1967:173). Thus not only trade but also the routes along which trade passed were of political concern. Furthermore, in the Eurasian steppes the caravan trade was basically run by professional merchants from sedentary countries, who specialized in this trade. The Great Silk Route was run by the Soghdians and later on Nestorians and Muslims from Middle Asia; the Volga Route was run by merchants from Khwarazm and the countries of the Caliphate. In their attempts to monopolize the trade they sometimes formed something akin to guilds and naturally made the greatest profit out of the trade. Caravan trade was always linked first and foremost to the political and economic situation in a wide outside world, which sometimes comprised several continents. Nomads could sometimes join in this trade, utilizing it in their own interests. But, as a rule, they neither created this trade, nor did they determine its development and its fortune. Even amongst the Tuareg 211

Modes of nomadic

adaptation

to the outside

world

profits from the caravan trade fluctuated considerably from century to century and from one group to another (Capot-Rey, 1953:303; Bovill, 1958; Keenan, 1977:34-5). The great geographical discoveries and shift of trade routes at the beginning of the modern period were a blow to intercontinental, and partly even to transcontinental caravan trade. In the long-term perspective caravels became more important than caravans (cf. Hames, 1979:381); just as the regular armies of sedentary states which were equipped with firearms began to gain victories over the irregular cavalries of nomads. However, in this respect cars were a much greater threat to caravans than caravels were. As a whole trade with the sedentary agricultural and urban world was not something from which nomads profited, but it was vitally necessary to them from the economic point of view. More often than not it was not profit which they derived from this trade, but the elementary means of existence. Submission and the different forms of the dependence of nomads on sedentary societies For nomads, and most particularly for nomadic aristocracies, not everything about incorporation in a sedentary state was negative. Cases are known where such incorporation was confined to more or less formal acknowledge­ ment of the state's sovereignty, frequently supplemented by an obligation to provide military service. In this case the position of nomads was basically determined by two factors: a simple alignment of their forces and those of the state, and the specific policy of the latter with regard to nomads. Such a situation, although it hampered nomads, nevertheless also involved certain benefits for them. Along with increasing opportunities for exchange and trade, they also received payment for military service. However, they never let pass an opportunity to pillage the sedentary population when the right situation for this arose, for example if a government was weak or turned a blind eye to such pillaging (see, for example, Lambton, 1953:139 on the situation under the Qajars; cf. Bates, 1971:123). Just as a dynasty which was nomadic by origin did not always put the interests of nomads before those of its sedentary subjects, so a sedentary government could not always effectively defend the interests of agricultural­ ists before those of nomads. The submission and encapsulation of nomads on the state level could be accompanied by their relative freedom on the local level. Nomads did not always manifest loyalty to a dynasty, even if this 1

1

In this section, with individual exceptions, the position of nomads in states in the postcoloninl period, where this period was already begun, will not be examined. There is already an extensive literature on this subject and the number of works being published on the theme in constantly growing.

212

Submission

and the different forms of

dependence

dynasty had itself emerged from amongst them. Even more difficult was it for a state to rely on nomads, quite apart from its sedentary inhabitants whose way of life, culture and frequently even ethnicity were different from those of nomads. Right up to the nineteenth century, and even to the twentieth century in some places (see, for example, Salzman, 1979:441-4 on the Shah Nawazi Baluch) this situation was particularly characteristic of the Near and Middle East, where states could not always effectively control the nomads who were more or less formally under their sovereignty. The medieval history of Iran and North Africa can serve as an example here. In Iran, from the eleventh century, nomads, in order to obtain and defend profitable and privileged positions both with regard to the state itself and to its various groups, strata and classes had to be powerful, that is centralized; this favoured the development of social stratification amongst them. A weak state (and the state in Iran in this period was weak more often than not) in its relations with the restless and recalcitrant nomadic element which, in effect, made up its basic mass of warriors, looked for a way out by incorporating the nomadic aristocracy into the ranks of the ruling class, bribing it with titles, official posts, gifts and lands. Thus in the fourteenth, fifteenth and sixteenth centuries, and sometimes even later, the chiefs of some nomadic tribes and their subdivisions were nominated rulers of certain provinces and districts and frequently, after receiving approval for this from their government, their duties became hereditary (Lambton, 1953:89f.; Petrushevsky, 1949:75f.; Istoriia Irana, 1977:173, 190). Furthermore, nomadic leaders were managers of those territories the revenues from which were marked out for one or other tribe. These territories included not only pastures, but also arable lands on which there lived peasants who were exploited. Admittedly, it should be borne in mind that in Iran, from the Saljuq period and most particularly in the period following the fall of the Hulagii's dynasty, there was no general opposition to speak of between any nomads, on the one hand, and sedentaries and the state, on the other. Different nomadic polities had different statuses in the various states on the territory of Iran, even in those states which had been created by nomads. Consequently, the leading positions in its own polity did not always guarantee the nomadic aristocracy a corresponding position in the state, if only because the state utilized the struggle for power between the different nomadic polities, setting off one group of nomads against others. The only exceptions to this rule occurred when aristocracies which were nomadic by descent lost their links with corresponding nomadic polities (Fernea, 1970; Salzman, 1974:208). Similar relations between nomads and states became most typical of North Africa alter the appearance there of the Banu Hilal nomads in the 213

Modes of nomadic adaptation

to the outside

world

eleventh century, and different variations of such relations continued to exist here for many more centuries, even in the periods of the strengthening and centralization of state power (for a more detailed discussion of this point see Chapter 5). 'Thus, by degrees, these Arabs, settled for scarcely more than threequarters of a century, came to constitute a political force, on occasion a dominant one, and were to remain such until the beginning of the twentieth century' (Julien, 1956a:118). But in North Africa the position of different nomadic units within the same states also varied very considerably. It goes without saying that situations such as those which have been described were, as a rule, very favourable for nomads; however, they were not so very rare, particularly in places where nomads and agriculturalists were not completely separated by ecological barriers. Otherwise the governments of sedentary states sometimes invited individual groups of nomads into their lands on account of internal political considerations and/or because they needed military assistance from nomads. This, in particular, was the policy of the Samanids and Qarakhanids in Middle Asia (Barthold, 1963:382; Barthold, 1963b:124). Admittedly, at the time of the Qarakhanid conquest the Turkic guard of Samanids turned out to be unreliable, but this experience was not put to any good use. When the Khwarazm-Shahs needed an army to pursue their conquests and strengthen their power in the lands they had conquered they drew nomadic Qipchaqs into their service and intermarried with the chiefs of the latter. Thus, under the Khwarazm-Shah Takish, a strong military estate was created. However, this estate still had the traditional shortcomings, the conse­ quences of which turned out to be fatal for Khwarazm. The Qipchaqs were very unreliable when fighting nomads, but as if to make up for this were ruthless in their plunder of sedentary territories which had been captured and set the local population up against the Khwarazm-Shah. Furthermore, they turned out to be inclined to court intrigues. In the reign of Khwarazm-Shah Muhammed they upheld his mother, Turkhan Khatun and were in opposition to the Shah himself (Barthold, 1963:413, 445). Muhammed clearly did not trust his army. Perhaps this was one reason behind the fatal decision to spread the army out over the towns of Maveraunhnar instead of massing it together to go into decisive battle with the advancing Mongols. Both Sasanian Iran and Byzantium, and Rome before this, hired Arabs to safeguard their borders from one another. In the Fatimid period the migrations of Bedouin tribes from Arabia into Egypt continued. The. Bedouins joined the army and thus became part of the ruling class. In Egypt the adverse effects of the Bedouins' presence were felt very strongly. Thus they were encouraged to migrate westwards, which was the beginning of the Hilal movement (Margais, 1931:78; Semenova, 1974:105). 214

Submission

and the different forms

of

dependence

Very different dynasties - the Ghaznavid, Timurid, Safavid and Mogul were keen to employ Afghan chiefs alongside their own troops and awarded them with lands for their services. A particularly large number of Afghans emigrated to India during the period of the Lodi (1450-1526) and Sur (154055) dynasties, in which they made up the ruling strata (Reisner, 1954:63). However, it is also known that there were quite a number of cases in which the status of nomads within a sedentary state was more ambiguous and not so beneficial for them, and especially not for ordinary nomads. The pressure of state power was always felt, if not directly then obliquely, if not always in reality, then its potential was always there. One of the most troublesome problems which nomads almost always faced consisted in taxation and other obligations which states tried to impose upon them. Even in medieval Iran nomads, especially those who did not come from privileged tribes were taxed by the state (and this was quite apart from the military service they were obliged to contribute), although at a lower rate than peasants (Petrushevsky, 1949:312f.). There were particularly sharp increases in these taxes after the reforms of Shah Abbas I and again in the reign of Nadir Shah. Frequently entry into the structure of a sedentary state was less painful for nomads when this state held sway over a large expanse of territory and successfully conducted a policy of conquest. For then economic benefits bought political submission, the pillaging of aliens compensated with interest for their inability to pillage 'their own' agriculturalists and townsmen. As has been noted by Watt (1953:20),'. . . the Qur'an appeared not in the atmosphere of the desert, but in that of high finance.' In a state in which Islam had recently emerged, the problem of the Bedouins was felt very acutely. Although Muhammad neither liked, nor trusted Bedouins (see, for example, Qur'an, Sura 49, verse 14), it was they who made possible the success of the state which he had created (cf. on the other hand Asad, 1973:65-6). However, Muhammad was farsighted enough not to try to change the basis of their social organization. From the Bedouins who, according to the cynical phrase of Caliph Umar, 'furnished Islam with its own material' (Hitti, 1956:29) and of whom a very considerable number had migrated from Arabia, all that was demanded was that they accepted Islam, that is that they acknowledged the state's sovereignty. Evidently, it was only the benefits from incipient conquests which guaranteed their loyalty to the new state and new religion. Arabia itself, from the time of the first Caliphs up to the appearance of Wahhabism in the eighteenth century, had almost no knowledge of a single power and a single statehood, and the Bedouins were fairly indifferent towards Islam. '. . . It is true that on the Turkish borders the Bedouins, out of prudence,

Modes of nomadic

adaptation

to the outside

world

keep up an appearance of being Muslims; but they are so lacking in reli­ gious rigour, so slack in their devotions, that in general they are regarded as infidels, knowing neither the law nor the prophets. Indeed they are quite willing to say that the religion of Mahomet was not made for them' (Volney, 1959:212; cf. Burckhardt, 1831:99f.). To a certain extent history repeated itself when Wahhabism emerged. It was primarily in the oases of Arabia that this teaching spread. But the notion of Jihad - a holy war against 'polytheism' - which was being developed by Wahhabism resulted in practice in raids and military expansion, and this was fairly attractive to Bedouins. If in the first years of the emergence of the Wahhabi state the Saudis were constantly at war with the Bedouins, from the eighties of the eighteenth century they began to work together more and more as allies (Vasilev, 1967:125,130f.), although a number of the imposed limitations and the obligation to pay zakat (tax legitimately paid by Moslems on their possessions) were clearly not to the Bedouins' liking. In essence the first state of the Saudis was created by the elite of the sedentary towns and oases in Najd in order to halt the endless internal wars which made normal conditions for production impossible, the struggle with the Turkish threat and common military expansion under the banner of a reformed Islam. The nomadic aristocracy only temporarily joined up with this elite, partly because of profits it could derive by doing so, and partly as the result of compulsion. The Bedouins were always ready to cast off the Saudi yoke, as the wars of Mohammed Ali which led to the complete destruction of the Saudi state showed. According to an interesting, although unproven suggestion of Lattimore (1967:222-3), the broad policy of conquest pursued by Tibet in the eighth century A . D . was conducted for the benefit of nomads, so as to capture their interest and draw them to the Lhasa state. An analogy between this situa­ tion and Muhammad and his Bedouins inevitably suggests itself here. Frequently a sedentary state, even when it was stronger and militarily had the upper hand, resorted to a policy of indirect government when dealing with nomads, not wishing directly to alter their sociopolitical organization, particularly in its lower levels; it was satisfied merely to encapsulate nomads within the limits of the state. As a result, sometimes 'administrative nomadic chiefdoms' of their own particular kind emerged. But such a policy in itself already led to more or less marked social transformations. The nomadic aristocracy was sometimes partially incorporated into the ruling classes. At the same time the help of the government and the functions of the mediators between the state and ordinary nomads in a nomadic polity strengthened the position of the nomadic aristocracy vis d vis the ordinary nomads (cf. Marcais, 1913:238ff.; Stenning, 1959:73-7; Barth, 1962:349; Irons, 1979:371; Salzman, 1974; Salzman, 1979:433, 434ff.). From the first century B.C., when part of the Hsiung-nu were forced to

Submission

and the different forms of

dependence

acknowledge the sovereignty of China, China conducted a policy of indirect rule with regard to the nomads she had subjected. This involved her in supporting and bribing local nomadic aristocracies, and sometimes even in directly interfering in the internal affairs of nomads, enabling their social organization to be transformed. Thus the Mongolian elite, the leaders of the aimaks and khoshuns which became the permanent territorial-administrative units under the Manchus, was included in the Ch'ing bureaucracy and received a salary from the government, while at the same time it exacted dues from ordinary nomads for its own use (Istoriia Mongolskoi Narodnoi Respubliki, 1967:187ff.; Ermachenko, 1974:159; Shirendyb, 1975:230; see also pp. 124,177-8 of this book). In Kam (Qam) the Manchu government tried to turn Tibetan nomad leaders into bureaucrats, making the ordinary nomad responsible for maintaining their leaders; at the same time the government exacted taxes from the nomads for its own use (Kozlov, 1947:172-3, 175). Even Russia at one time experimented with a policy of indirect rule with regard to the Kazakhs and Kalmucks until she consciously abandoned this policy in the nineteenth century. However, with regard to the aristocracy of different nomadic peoples, and even amongst the same peoples, the policy was different in different historical periods, varying according to the corresponding political situation. Amongst the Kazakhs, the influence of the 'white bone' was at first supported and consolidated by the Russian administration, but then undermined and abolished. In the 1867-8 reforms the Jenghizids were stripped of their estate privileges and put on the same footing as 'countrymen'. On the other hand, in the eighteenth century the Kalmuck aristocracy was definitively separated into a closed and privileged estate (Nebolsin, 1852:7-9; Palmov, 1929:18-19; Erdniev, 1970:206-11). Kalmuck noyons, the heads of individual uluses had the right to inflict physical punishment on ordinary nomads and imposed an annual tax on them on a scale of one head in ten of stock (Pallas, 1809:485). The Khiva Khans supported the Turkmenian chiefs they had subjected and enabled them to seize lands for themselves, part of the taxes from which the Turkmenian leaders were allowed to collect; the Russian government enabled property-inequality to develop amongst the Turkmen, something which was partly linked with their sedentarization (Markov, 1976:231-2). In Iran a policy of indirect rule was practised by the government even in the twentieth century (Salzman, 1974:204-5; Salzman, 1978b:544-5; Fazel, 1973:135). In the sixteenth and seventeenth centuries the Moguls and Safavids collected tributes from the Afghans and made use of the services of their leaders. In return they made it possible for the latter to consolidate their 217

Modes of nomadic

adaptation

to the outside

world

power, released them from taxation and granted them deeds stating that they owned lands, etc. (Reisner, 1954:55, 225). It was to the Safavids, on account of his elevation by them, that Malik Sado Abdali of the Saddozais, from whom the founder of Durrani power was later descended, had obligations. The Durrani Shahs pursued the same policy. In the majority of tribes the elections of khans were controlled by the Shah, and the more obedient a given tribe was to the Shah, the more obedient it was then to its Khan(Elphinstone, 1819, i:255,264). 'It may be worth while to remark the circumstances on which the extent of the power of the chiefs depends. In tribes that are obedient to the King, the Khann derives much influence from his employment of collecting the royal revenue and raising the militia, and indirectly, from the emoluments attached to those duties. His personal character materially affects his power and influence' (Elphinstone, 1819, i:260-l). The power of Bedouin sheikhs also increased when they were supported by central powers (Awad, 1962:335; Marx, 1977:350). Frequently the sheikhs would act as mediators between their nomadic polity and those powers (see, for example, Burckhardt, 1831:96), while the latter did all they possibly could to draw the sheikhs onto their side and buy their loyalty. The Turks in Libya were limited to receiving taxes, not always regularly, which were collected by Bedouin sheikhs who kept part of them for their own use (Evans-Pritchard, 1949:96; cf. also pp. 282f. of this book about the situation in North Africa in the Middle Ages). In this way, the entry of nomads into a sedentary state, even in those cases in which they retained a considerable degree of autonomy, frequently enabled social differentiation to increase amongst them (cf. Irons, 1971:155; Salzman, 1974:209). However, a policy of indirect rule with regard to nomads was not always sufficiently reliable from the point of view of a state and its government. The dual loyalty of nomadic leaders, to their unit and to the state, if the appropriate conditions arose, could easily turn into disloyalty to the state. The histories of China, Iran and many other countries are sufficient proof of this fact. Nomads were all the more aware of the strong arm of the state when the latter adopted a policy of direct government towards them, or when it no longer needed to give them a special status, or when a state simply sought to liquidate the problem of nomads within its borders. In the course of history different Chinese governments frequently resorted with regard to subjected nomads to a policy of forcible assimilation in its most extreme manifestations, forcing them into marriage with Chinese. One early example of this is provided by the Uighurs in the province of Shensi in the twelfth century (Song mo jiwen, pp. 6-8, quoted in Maliavkin, 1974:92; see also Maliavkin, 1975:80-1). Later on, during the 21K

Submission

and the different forms of

dependence

Ming Dynasty, marriages between the Mongols who had remained in China after the destruction of the Yuan state were forbidden. In a 1372 decree of the Emperor Zhu Yuan-zhang it is written: 'The Mongols must not be let go back to their homeland. They must become related to the Chinese by the path of marriage and marriage between Mongols must be forbidden. Those who violate this law, men and women, will be turned into slaves' (Ming ji jie fu-li, quoted in Dalai, 1975:202). Much earlier, the Emperor T'o-pa Hong of the state of North Wei, himself a nomad by descent, adopted a policy of making his state as Chinese as possible, and encouraged mixed Hsien-piChinese marriages (Kriukov, Maliavin, Sofronov, 1979:94). In Tibet the tribal structure of many nomadic groups was destroyed by pressure from the Lhasa government (Kozlov, 1947:35, 211; Ekvall, 1968:29; Reshetov and Iakovlev, 1975:210). At the same time as nomads felt themselves to be fairly free in a weak Iran, in neighbouring Turkey the government obstructed the creation of large nomadic polities and, disregarding the economic interests of nomads, pursued a policy of controlling their pastoral migrations, and limiting their autonomy and internal integration. The uprisings of the Kurds in the nineteenth and twentieth centuries were only a belated reaction to such a policy (de Planhol, 1969:83f.). The Fulani, who constituted a minority in the Muslim states of West Sudan in the period preceding the Uthman dan Fodio movement, had to conform to the policies of these states, participate in their wars, pay taxes and even give up their women to local rulers (Stenning, 1959:6; Stenning, 1965:365, 367; Horton, 1971:111). It usually happens that the more the balance of power changes, the worse the position of nomads in a sedentary state becomes. Their obligations increase. Even in Afghanistan, where for a number of reasons the position of nomads until the recent past was better than it was in many other countries, nomads had, willingly and unwillingly, grown accustomed to paying taxes in order to avoid conflicts with the government (Schurmann, 1962:317, 322). Frequently a disintegration of the higher levels of the sociopolitical organization of nomads occurs and the significance of property-inequalities increases, and traditional institutions of reciprocity and redistribution cease to function (see, for example, Digard, 1973:48 on the Bakhtiari). The nomadic economy is integrated into the local, or even into a wider economic system and often is compelled to adapt itself to the sedentary-agricultural sector (as the leading sector) of this system in conditions dictated by the latter. In the present day this tendency can be regarded as the predominant one, although in certain areas its predomi­ nance was established much earlier. Wherever nomads do not occupy the predominant position in a particular ecological zone, even their pastoral migrations are sometimes dictated by 219

Modes of nomadic

adaptation

to the outside

world

the agrarian productive cycle of the villages through which they move, and sometimes they do not even own the pastures. In Russia the best Kazakh, Kirghiz, and partly also Kalmuck and Buryat lands in the nineteenth century were taken for the use of the Russian immigrants (see, for example, Shakhmatov, 1964). The cycle of migrations of the Turkish Yoruks is dictated by the landowners, in accordance with the requirements of their own productive rhythm. The Yoruks do not have their own pastures and so have to pay agriculturalists or peasant communities for the use of pastures (Bates, 1971:124-7). One result of these changes is that the pastoral nomadic economy loses still more of its subsistence-oriented features, and is drawn more and more into exchange and trade. Different forms of trading are practised in local markets, and they include mediatory, barter and money trade, with due regard for fluctuations in market prices (see, for example, Barth, 1962: 346-7; Barth, 1964a:72; Barth, 1976:74; Ferdinand, 1969:143; Fazel, 1973:133, 134; Swidler, 1973:28; Bates and Lees, 1977 on the nomads of the Middle East). Such a situation may result amongst nomads in their hiring themselves out for work in sedentary agricultural and urban sectors of the economy (see, for example, Capot-Rey, 1953:166; Keenan, 1977:173f. on the nomads of the Sahara; Frantz, 1978:104 on the Fulani; Swift, 1979 on the Somali; Dahl and Hjort, 1979 on the pastoralists of East Africa; Pastner, 1971:175; Salzman, 1971:187; Swidler, 1973:32; Balland and Kieffer, 1979:81 on the nomads of the Middle East). This situation enables nomads to acquire the goods they need, but in conditions which are extremely varied. In the time of the Manchu dynasty trade became a direct weapon with which Chinese merchants could exploit the Mongols, a practice in which they were assisted by the Mongol aristocracy. The latter utilized the collec­ tive responsibility of the groups they governed and guaranteed Chinese mer­ chants the payment of duties and percentages, and sometimes the Mongol aristocracy even put capital into Chinese firms (Grumm-Grzhimailo, 1926, m, pt. 1:353-4; Maisky, 1959:167ff.; Lattimore, 1967:201-6; Istoriia Mongolskoi Narodnoi Respubliki, 1967:201-6; Burdukov, 1969). However, other cases are known in which a natural peasant economy turns out to be less adaptable to a money-market economy than a nomadic one is. Thus Afghan nomads sold goods to the Hazara in remote areas and deferred payment for these goods for a year, but at an interest of 100% per annum, and anyway they charged very inflated prices. They also lent money under the same conditions. As a result much land in Hazarajat passed into the hands of nomads (Ferdinand, 1962:130f.; Ferdinand, 1969:149; Glatzer, 1981; see also p. 158 of this book on other nomads in the Middle East). However, it is important to make the point that it is basically rich 220

Submission

and the different forms of

dependence

stockowners who derive benefits from such trade-monetary relations with agriculturalists. One other policy which sedentary states have frequently pursued and still do pursue with the nomads under their sovereignty is sedentarization over a fairly wide spectrum, ranging from voluntary, although encouraged sedentarization, to the compulsory and forcible. It is no coincidence that nomadic subjects in a sedentary state have sometimes tried to remain nomads so as to avoid tax demands and pressures from the state powers. In North Africa the Numidian king Masanasses had already in the second century B.C. begun to pursue a policy of mass compulsory sedentarization. The Romans continued this policy (Capot-Rey, 1953:185, 282; Julien, 1956:97f., 133f.; Julien, 1956a:307; Bousquet, 1957:32), and Strabo (Strabo, XVII.3.15) called it the turning of nomads into social people. It seems extraordinary that the attitude of civilized sedentaries to nomads and their problems has changed little in the last two thousand years. In all fairness, however, it must be pointed out that the sedentarization of nomads in North Africa in antiquity was accompanied by the construction of large irrigation works, the building of roads, etc. But later on it was far from the case that sedentarization was carried out with so much forethought (on the policy of sedentarization see, for example, Pershits, 1961:210ff.; Capot-Rey, 1962:307; Keenan, 1977:173ff. on the Near East; Reisner, 1954:55; Bates, 1971:121; Ferdinand, 1962:47-9 on the Middle East; Trubetskoi, 1977, for a general summary of contemporary data). For the reasons which have been mentioned it was not always successful and effective, and rarely was it beneficial for nomads. Nomads willingly became landowners, but were considerably less enthusiastic about being agriculturalists. Even today in many countries it is only the rich and elite nomads who are in a position to acquire land and then become partially or fully sedentary (see, for example, Lewis, 1961:93 on the Somali; de Planhol, 1959:529-30; Bates, 1971:121 on Turkey; Rosenfeld, 1965:77, 79; Stein, 1967:102ff. on Arabia; Stauffer, 1965:294; Fazel, 1973:133, 139; Digard, 1973:50 on Iran; Glatzer, 1981 on Afghanistan; see also pp. 158-9 of this book). As a whole the acknowledgement of the sovereignty of sedentary governments sometimes benefited separate groups of nomads and/or their ruling strata. But real submission was linked to loss of political indepen­ dence, the jeopardizing of traditional institutions and forms of sociopolitical organization, direct or indirect exploitation by the state and its ruling classes, and wherever there were ethnic differences with the threat of assimilation. Almost always submission was a path which nomads had to take, but they did so involuntarily, and under pressure. It is no coincidence that this form of submission is most typical of the modern period. 221

Modes of nomadic

adaptation

to the outside

world

Subjugation and the different forms of the dependence of sedentary societies on nomads What we are dealing with here is the wide spectrum of turbulent interrelations between nomads and the sedentary world which appeared as a result of advantages contained in certain aspects of the nomadic way of life mobility and military superiority - the latter particularly in places where nomads breed riding animals (Pershits, 1976; Bonte, 1977b:44). The spectrum of these relations varies from irregular raids on and robbery of agriculturalists and townsmen to the imposition on them of more or less long-term relations of protection and dependence. The benefits from such non-economic relations for nomads are so evident that they need not be examined in detail here. Nomads are in a position in which they are able to acquire the agricultural products and handicraft goods they need (and also livestock) by force, or by threatening such, while giving little or nothing in exchange. Thus wherever nomads have the corresponding opportunities, their raids and pillaging become a permanent fixture. For many nomads they were an important supplementary means of livelihood. Thus, according to Burckhardt (1831:57, 107-8, 234ff.), many groups of Bedouins in Arabia were unable to feed themselves on the proceeds of their pastoral economy and so had to become involved in robbery. Spooner (1975:178) notes that amongst the Baluch raids have become a sup­ plementary form of economic activity. In practice, nomadic raids have been recorded in all regions where nomadism is widespread (see, for example, Klima, 1970:110 on the Barabaig; Leeds, 1965:109 on the Chukchi; Kozlov, 1947:214, 305; Ekvall, 1968:40, 52-3 on the nomads of Tibet). Correspond­ ing data about the nomads of the Eurasian steppes, Near and Middle East is too voluminous and easily available for there to be any real point in my making a selective list of it here. Until the modern period with its revolution in things military, sedentary states were unable to find any permanent solution to the military problem of how to defend themselves against the raids of nomads. It is this fact which explains the following paradox: in different regions and in different historical periods sedentary states built expensive defensive systems on their borders with the nomadic world, but again and again nomads have shown that these systems were ineffective (see, for example, the Great Wall of China, or the Roman limes in North Africa, or the Darband Wall and other walls which the Sassanids built against invasions from the north, or the long walls built in the eighth century by the Arabs to defend the oases of 2

2

Such relations can exist both between different nomadic groups and polities, and also within the latter. The significance and consequences of these relations have been briefly examined In chapter 3 - see pp. 162-4.

222

Subjugation

and the different forms of

dependence

Maveraunhnar, or the whole system of fortresses and fortified points set up by the Kievan prince, Vladimir). Only the Romans were able effectively to protect their agricultural and urban territories in North Africa for a few centuries, and even to push their border southwards. But in the end this border was undermined. The Russians were able to expand their borders only in the seventeenth century. Nevertheless, raids were essentially too unreliable and unpredictable, too dependent on the balance of power between nomads and the sedentary population at each concrete historical moment. When both sides, nomads and the sedentary population, preserve their political independence such exploitation cannot be guaranteed as reliable, does not have fixed forms and risks destroying the very foundations of the economic life of the sedentary population (see, for example, Capot-Rey, 1953:174; de Planhol, 1966:280 on the Sahara). In the long term it was more beneficial and safer for nomads to insist the sedentary population paid them regular tributes in return for their protection and defence against other groups of nomads, and just so as to be left in peace. On the local level such relations were most widespread in the Near East (see, for example, Musil, 1918:42; Pershits, 1961:133f.; Awad, 1962:335; Capot-Rey, 1962:304; Rosenfeld, 1965:77; Stewart, 1973a: 381; Coon, 1976:195-6), and partly also in the Middle East. They were considerably less common in the Eurasian steppes. However, long-term and stable relations of the tribute type between nomads and the sedentary population which were established on a local level and, as a rule, as the result of a historically shaped alignment of forces, rather than a single mass conquest, were basically characteristic of those regions and periods in which there was no sedentary state at all, or else it was very weak. It is no coincidence that such relations have become most widespread in Arabia and particularly in the Sahara, or in Baluchistan in the Middle East where the type of sedentary life, concentrated as it is in small oases which are separated by wide expanses of desert, is doomed to military weakness. In other cases the state has not always been able to protect its subjects, but at least warded off strong dependency on nomads. Wherever a sedentary state opposed nomads as a real force with which they had to contend, nomads had to use force to acquire the products they needed involving either the subjugation or the conquest of the state in question. All ways in which nomads subjugate or conquer sedentary societies can be regarded as a form of political adaptation to the outside sedentary world. They all involve a transformation of the sociopolitical backwardness of nomads into a military advantage, that is a strong, mass and mobile military organization united by (he prospect of exploiting an agricultural and urban population. This guarantees nomads an uninterrupted and unhampered 223

Modes of nomadic

adaptation

to the outside

world

supply of agricultural and handicraft products on the non-economic basis which is most profitable to them. These supplies are acquired in various ways. The most typical and widespread are the following: (1) Direct irregular and uncontrolled pillage. If the conquest of sedentary territories is not a short-term episode and nomads wish to further establish their supremacy over the conquered population, such pillaging cannot last long. Amongst the nomad leaders there are always those who realize that the goose which lays the golden egg should not be slain. Admittedly, they are not always able to act this way at the apposite moment. When the Khitan leader, Te-guang, in the tenth century simply annexed conquered China to his own state calling her Great Liao, he was unable to pursue the exploitation of the conquered country in any organized form. General uprisings followed and the newly-made emperor was forced to return home. In his explanation of why this state was so short lived, Te-guang showed his understanding of the meaning of events when he admitted to three mistakes he had made: he had imposed excessive demands on the resources of the Chinese population, he had allowed the Khitans (K'i-tan) to pillage, and he had not appointed local rulers in time (Taskin, 1975:98). Subsequently, the Jiirchens, Mongols (although not immediately) and Manchus could have made considerable use of the unsuccessful experiences of the Khitans. (2) Tribute. Tribute is taken to mean the external form of dependence and exploitation whereby conquered groups give up part of the goods they produce and/or fulfil other duties for their conqueror, while at the same time preserving their economic and sociopolitical organization, although not always in a complete and unchanged form. Usually they are not integrated (or not fully integrated) into the sociopolitical organization of the ruling group. As a rule, the origin of tribute relations is closely connected with non-institutionalized military pillage and contributions and it is a collective, rather than individual, form of dependence (Khazanov, 1972:168-9; Khazanov, 1975:122-3; Khazanov, 1975:159; Khazanov, 1979; Pershits, 1979). Different forms and different levels of tribute-relations between nomads and sedentary populations are known. In some cases, which have already been noted, the population of a village in the Middle East or in an oasis in the Near East would pay tribute to a specific nomadic group or unit, or to separate nomadic aristocrats for 'patronage' and protection from other nomads, and even more so in order to buy off the raids of their actual patrons (see, for example, Lambton, 1953:160; Pershits, 1961:1331:; Rosenfeld, 1965:78, 79; Coon, 1976:195). In other cases a sedentary stale paid tribute to a nomadic one (for example, China to the Hsiung-nu, or the Russian princedoms to the Golden Horde). (3) Direct taxation. In contrast to tribute, this is a more developed and 224

Subjugation

and the different forms of

dependence

centralized form of exploitation and is only possible in a situation in which nomads occupying the dominant position in a state, which they have created after conquering sedentary countries, have the necessary fiscal apparatus to hand. As a rule, this apparatus is basically staffed by the bureaucracy or other individuals of the sedentary population who go into the service of their conquerors. In such cases extremely predatory farming forms of taxation are frequently practised at first. Only subsequently were these methods sometimes changed by fixed taxation. This was often to a considerable extent due to the work of such distinguished statesmen as Yehlu Ch'uts'ai and Rashid al-Din, who had emerged from amongst the subjugated sedentary population and for whom the word realists is far more suitable an appellation than collaborationists. At any rate they understood that politics is the art of the possible, although the 'possible' could cost them their heads. At the same time the work of such farsighted nomadic rulers as Ogodei and Ghazan Khan should not be underestimated. But as a general rule, such a situation arose relatively rarely, and certainly not after every conquest of a sedentary state by nomads (see Chapter 5). In other cases a similar situation can arise when groups, which are nomadic or pastoral by descent, occupying the dominant position in a state at the same time become integrated with the agriculturalists in one economic system based on a social division of labour. This has happened in the Interlacustrine states in Africa. I examine these cases in more detail further on in the book. (4) The creation of agricultural and handicraft sectors of economy within nomadic society itself. The only cases which are relevant here are those which are directly or indirectly linked with the submission of sedentary groups or societies, or their short-term or long-term consequences, and not with the processes which take place within nomadic societies themselves (for example, the sedentarization of impoverished nomads). The benefits nomads derive from such sectors of economy are evident. They make the economy of nomads more diversified and, if they do not eliminate this entirely, they lessen the dependence of the economy on the outside world. In East Africa hunting castes and ethnic groups in pastoral societies play a similar role (Monod, 1975:141; Dahl, 1979:262, 267; Torry, 1979:520), but as yet this has been studied in little detail. However, methods for creating such sectors of economy can be different. In some cases nomads have simply moved craftsmen and agriculturalists from sedentary countries they have conquered onto their territories, or else they have used captured slaves in such a capacity. Such a practice was very widespread amongst the nomads of the Eurasian steppes, and also in the Near and Middle East (see, for example, Capot-Rey, 1953:168-70; Pershits, 1961:138; Awad, 1962:335; Johnson, 1969:151-2,156). In the first centuries

Modes of nomadic

adaptation

to the outside

world

of the Hijra the Arabs settled in the Hijaz agriculturists from conquered countries, whom they made into slaves (Al-Ali, 1959:252). In other cases nomads attracted craftsmen and peasants into their territories, because the former were interested in their production. In such cases the position of those craftsmen and peasants was easier, although they were still dependent on and exploited by the nomads. Ahmad Shah Durrani forced Indian and Iranian merchants and craftsmen to settle in Qandahar. Practically none of the craftsmen in his state were Afghans (Reisner, 1954:82). But at the end of the twelfth century and the beginning of the thirteenth Chinese craftsmen lived amongst the Kirghiz, having agreed to do so willingly (Barthold, 1963e:505). In ancient times the Chinese stated that '. . . the bondmen and bondwomen of the borderdwellers lament their arduous life, amongst them many want to run away, and they say: "There are rumours that life is placid amongst the Hsiungnu, but what can you do when severe guards have been set up?" Despite this sometimes people do nevertheless run off beyond the fortified border' (Pan Ku Hanshu, Chapter 94b, quoted in Taskin, 1973:41). In other cases, in the Near and Middle East, for example, nomads, agriculturalists and craftsmen are to be found within the boundaries of the same tribal units in which nomads occupy the dominant position (see, for example, de Planhol, 1966; Barth, 1973:20). Sometimes such dependent groups, especially those consisting of craftsmen, are regarded as an inseparable, component and specialized part of the nomadic society proper which remains dependent simply because it has nowhere else to go. However, by origin these groups are nevertheless usually linked not with economic specialization within the nomadic society in question, but with the submission of sedentary groups, or with a general situation created by nomadic invasions and conquests when acceptance of a dependent position on and patronage from nomads was their only option (see, for example, Capot-Rey, 1953:226; Nicolaisen, 1963:18 on the Tuareg; Musil, 1928:281; Rosenfeld, 1965:77; Coon, 1976:196, 200 on the Bedouin of Arabia; de Planhol, 1966:273f. on the nomads of the Near and Middle East). (5) The seizure of landed property and receipt from it of dues in the form of rents, direct exploitation of peasants who are turned into tenants, metayers, etc. As we have already seen this is possible even in those cases in which nomads are incorporated in a sedentary state. Conquest and submission, however, present the best opportunities for this form of exploitation (see, for example, Philby, 1922, n:13; Pershits, 1961:137; Rosenfeld, 1965:79 on Arabia; Barth, 1962:346 on Southwest Asia; Reisner, 1954:110-11, 146, 225; Schurmann, 1962:263-^; Ferdinand, 1969:144 on Afghanistan). Of course, there are different variants, including more or less centralized ones, of each of the forms of dependence and exploitation which have been 226

Subjugation

and the different forms of

dependence

examined. And, it goes without saying, the existence of one of these forms in a given situation does not preclude the existence of other forms there. At the same time the predominance of one or other forms depends on many different factors and on the combination of such - the general historical situation, economic and sociopolitical characteristics of both conquerors and, particularly, conquered, the nature of the submission and conquest and, finally, the relations established between conquerors and conquered and the ecological background against which these relations take place. 'This event, the sparks from which have flown [in all directions] and the harm from which has spread to everyone; it spread through the villages like a cloud blown by the wind', wrote Ibn al-Athir of the Mongol invasion (quoted in Tizenghausen, 1884:2). There are numerous examples in history of the conquest by nomads of sedentary societies and states, many of which have been of world-wide historical significance. Apart from the emotional overtones of reactions such as the one quoted above, scholars from the time of Ibn Khaldun, if not the Jewish prophets, have been assessing the influence that such invasions have on the functioning and development of sedentary societies. Considerably less attention has been paid to the ways they have affected nomads. Meanwhile the subjugation and conquest of sedentary populations, particularly large-scale ones, in the majority of cases have led to a quick and marked transformation in the sociopolitical organization of nomads, in many cases considerably more marked a transformation than occurs amongst the agriculturalists and townsmen they have conquered. The result of this transformation has been the emergence of so-called nomadic states.

227

5 Nomads and the state

Nomadic statehood and the conditions of its emergence The terms 'nomadic statehood' and 'nomadic state' are extremely vague and imprecise. Strictly speaking, only such independent social organisms which have reached an appropriate evolutionary level, have an appropriate political organization and in which the majority of the population consists of nomads, separated into ruling and subordinate strata and/or classes, can be called nomadic states. However, in a wider sense, those states which were created by nomads as a result of their conquest of other societies, particularly those in which nomads occupied the dominant position within them, frequently have been called nomadic states. Finally, states with nomadic dynasties, or with dynasties which are nomadic by origin have existed. Furthermore, in certain states nomads or pastoralists have occupied the dominant or privileged position for a number of reasons not directly connected with their conquest. The reason these terms are used widely is partly due to the fact that the states to which they are applied are sometimes genetically linked to one another. Thus, for the time being I shall use the terms 'nomadic statehood' and 'nomadic state' in their wide sense and return to defining them more specifically in the final part of the chapter, after I have examined different examples of their emergence and the ensuing consequences of such. It is from conclusions drawn in previous chapters that I begin to examine the actual question of nomadic statehood. From the economic point of view pastoral nomadism is non-autarkic, and whilst pastoralism remains nomadic it needs the economic resources of the sedentary, primarily the agricultural and urban world. In view of this fact, non-economic ways of acquiring products they were unable to produce themselves, particularly the subjugation of sedentary groups, societies and states, were the most profitable for nomads. Conquest was only one of the ways submission could be brought about, and only if brought to its logical conclusion. A number of circumstances, about which a great deal has been written in 228

Nomadic

statehood

and the conditions

of its

emergence

previous chapters, limited the development of social differentiation and the emergence of the state amongst nomads suigeneris, that is only as a result of their internal development. More often than not the emergence of a nomadic state was directly linked to conquests, as a rule, of sedentary countries and regions. Admittedly, defence needs evidently could also cause a state or statelike polities to emerge amongst nomads. But in this case they were usually short-lived and extremely unstable. Moreover, amongst nomads there was always only one step between defence, and attack and expansion. Therefore, the subjugation and conquest of sedentary agricultural and urban societies is a specific manifestation of the external factor which, together with the internal, determined the particularities of the sociopoliti­ cal organization of nomadic societies and the general level of their development. In so far as nomadism is not only economic adaptation to the environment, but also an active adaptation to economic, sociopolitical and cultural conditions in the outside world, subjugation and conquest were an extreme method of adaptation of the second type. Adherents of the conquest theory of the state (Ueberlagerungstheorie) have usually cited and still cite facts about the conquest of agriculturalists by nomads or pastoralists in order to substantiate their views (see, for example, Gumplowicz, 1899; Oppenheimer, 1926; Thurnwald, 1935; Westermann, 1952; Eberhardt, 1952; cf. Pritsak, 1952:51). In my view the question is a far more complex one. First of all, two generalizations need to be distinguished carefully. (1) The emergence of nomadic states is linked, if not in all, in the majority of cases, with successful external expansion, primarily at the expense of a sedentary population. (2) By no means all non-nomadic states emerge as the result of conquest. In this way conquest is in general only one of the specific ways by which a state can emerge, but it is certainly not the only way (Khazanov, 1978; Khazanov, 1979). But if it is true that not every state emerges as the result of conquest, then I lie reverse assertion, that not every subjugation and conquest automatically entails the emergence of a state, is also true. For a state to emerge, subjugation and conquest must be sufficiently stable and large-scale, meaning that numbers and economic potential in the sedentary population must be sufficiently large. Apart from this, there must already be a certain degree of social differentiation in the societies of both the conquerors and the conquered. It is doubtful, for example, whether the conquest of one egalitarian society by another anyway is possible, and a prolonged and stable conquest in this case is even more doubtful. The very minimum of social differentiation can itself be insufficient to induce separately the emergence of the state amongst the conquered and even more so amongst conquerors without any actual conquest. But the result of subjugation or conquest, which act as catalysts, can be the

Nomads

and the state

emergence of superstratification (see, for example, Fried, 1967:232; cf. Wittfogel, 1963:324-5) which, on the one hand, fully or partially turns ethnic differences into social ones and, on the other hand, leads to the emergence of the state. Nevertheless, nomadic states have usually emerged as a result of the conquest of already existing states and/or sedentary territories, the urban and peasant population of which have already been separated into different strata and classes. Sometimes it is easier, when constructing a building, to use materials from the ruins of other buildings than to start completely from scratch. Sometimes, when only the upper storeys have been destroyed, it is easier to build them up again and repair. New buildings which are constructed or finished off in this way may be less beautiful and well proportioned than the preceding ones, but at least they are put up far more quickly. Thus, frequently, the states created by nomads made considerable use of the heritage they took on from their sedentary, and partly also nomadic predecessors. I shall try to elucidate below what exactly this meant. Irrespective of circumstances, a nomadic society had to be sufficiently consolidated before it could conduct a policy of broad external expansion. All or the majority of nomads had to derive certain profits from the expansion (or, to put it more generally, military successes), or to expect to derive profits in the near future. At the same time the stronger the objective of expansion, the greater was the need for the nomads in question to be centralized and united. However, it certainly is not always the case that increase in social differentiation amongst nomads had to precede external expansion. On the contrary, it might happen during the course of expansion and according to the success of expansion. Successful expansion presented nomadic society with fresh problems. It could promote the stabilization of political power, the strengthening of the social and property positions of the nomadic aristocracy, which had existed earlier or had once again emerged, transforming this aristocracy into the ruling class with regard to the subjugated population and a privileged leading estate with regard to ordinary nomads. Moreover, in specific circumstances, it opened the way for the transformation of the nomadic aristocracy into a ruling class over ordinary nomads. Such expansion could alleviate and solve the internal problems amongst nomads which emerged as a result of growing social differences, because it was able to carry these problems outside and solve them at the expense of the subjugated or conquered society. Nevertheless, the social and economic consequences of expansion could lead to a nomadic society losing just those of its characteristics which had enabled it to expand successfully. Then other nomads could set out on the same path as their predecessors had taken amd partially repeat the same process. Different nomads had different ways of solving the fresh problems 230

Types and tendencies

of the emergence of nomadic

statehood

confronting them as a result of successful expansion. But the solution chosen had a strong influence on the specific character and fortunes of particular nomadic states. The nature of the states which emerged as a result of subjugation or conquest by nomads of agriculturalists was to a great extent determined by the socioeconomic relations established between conquerors and con­ quered. In an analysis of these relations due account must be taken of the ecological factor. Three basic types of states which have emerged as the result of such conquest may be singled out. Each one of them is linked with specific paths of a specific evolution. Admittedly, it was rare for these evolutionary paths to be fully realized, and sometimes they were reversible. Consequently, when nomadic states are being examined, it is frequently preferable to talk of the corresponding peculiarities, tendencies and directions of their emergence, functioning and transformation. In practice these tendencies were frequently manifested at the same time within the same states. However, only one of them, as a rule, was predominant in a state in a specific historical period. But, of course, it should be borne in mind that every specific case is unique and unrepeatable. Nevertheless, a certain amount of schematization of historical facts is something which is unavoidable in any generalization. Main types and tendencies of the emergence and evolution of nomadic statehood States of the first type (or states in which the first tendency is predominant) are those in which the subjugation and conquest of the sedentary population basically result in vassal-tribute or other primitive, and not always com­ pletely regulated forms of collective dependence and exploitation. These forms were usually most stable and long-term where nomads and the seden­ tary population continued to inhabit separate ecological zones. Sometimes in this case the sedentary population would preserve its own state, with only a vassal dependence on the nomadic state; sometimes nomads and sedentaries were joined within one and the same state. Under all circumstances it was primarily in the political sphere that their very limited integration took place. Nomads and sedentaries lived side by side, but not together. The exploitation of the latter, sometimes predatory, did not touch the social and economic foundations of its society, nor did it entail (he emergence of a single socioeconomic system in addition to the political system. When a vassal dependence of a sedentary state on a nomadic one was established and especially in the cases in which nomads, agriculturalists and townsmen were incorporated in a single state changes were always

Nomads

and the state

stimulated in the nomadic society. The emergence of a nomadic state was linked with at least limited sedentarization. Thus, a nomadic aristocracy could not do without towns which Were the centres of political power, handicrafts and trade. Sometimes it was the aristocracy which caused these towns to emerge. Admittedly, the emergence of these towns looks somewhat artificial. It was not so much the state which existed on their account as that they existed on account of the state. Furthermore, in the states here being examined, such phenomena were not irreversible. All that they were able to do was to sometimes act as preconditions for further changes, but already on a different basis. Whilst external expansion continues and until the aggressive onslaught subsides, the first tendency frequently tends to predominate in the nomadic state. Afterwards, because of the changing alignment of forces, or for other reasons, far fewer opportunities arise for the primitive exploitation of the sedentary population, or else they cease altogether and three basic variants of further changes may be observed. The first is linked with the breaking away of sedentary territories or states, a decrease in the importance of agriculture and handicrafts and the decline of towns within the nomadic society, its sociopolitical structure once again becoming fairly primitive, and with the disintegration and downfall of the nomadic state. The second variant is caused by further development towards and final success in the sedentarizing process. As a result the nomadic society as a whole ceased to be nomadic and became an agricultural-urban society par excellence. Its nomadic structure to a considerable extent could still be preserved, but it developed in a completely different way. The third variant is linked to cases in which a nomadic state of the first type is transformed into a state of the second type. I shall now describe this variant. In nomadic states of the second type nomads, agriculturalists and townsmen were integrated into a single sociopolitical and, partly even, economic system. Such integration in a complete and finished form is relatively rare. Usually it affects only certain strata and classes in both sedentary and, more particularly, nomadic societies. (It is in these situations particularly that it is preferable to talk of a corresponding tendency in the functioning and transformation of a nomadic state rather than a specific type of nomadic state.) States of the second type are particularly characteristic of those situations in which nomads, after conquering a sedentary state, or during the process of conquest, moved onto the territory of this state and began to divide the same ecological zones between themselves and agriculturalists. In a state such as this a synthesis was taking place between the relatively less developed social relations of the conquerors and the more developed 232

The Eurasian steppes, semi-deserts

and deserts

relations of the conquered. Correspondingly, changes in states of the second type (or in states in which the second tendency of development predominated) occurred faster than they did in states of the first type. States of the third type are characterized by their having a single socioeconomic and political system, at the basis of which there is division of labour, as a rule between pastoralists and agriculturalists. In this case social stratification and the formation of classes developed along lines which for the most part coincided with economic specialization and ethnic differences. I do not think that the types and tendencies for the emergence and transformation of nomadic states which are here singled out are exhaustive. Certain others are known. For example, amongst nomads the state can emerge on account of internal sedentarization and/or the creation of agricultural sectors of economy in their society without any conquests, as a result of a religious movement (the Sanusis, for example). Furthermore, all the reservations I expressed about the conditionality of different typologies in preceding chapters are fully applicable in this instance. Such are a few preliminary ideas about the emergence and transformation of nomadic states. In order to describe these ideas more specifically it is necessary to examine the individual characteristics of nomadic states in the different regions in which nomadism is to be found. Unfortunately, the size of this book and the limitations of my knowledge only permit me to deal with individual examples in each region. I have tried to select those examples which, in my view, best illustrate the aims stated above. But, of course, the fragmentariness of the picture can only have an adverse effect on the final conclusions. The Eurasian steppes, semi-deserts and deserts Two ecological characteristics of this region which have influenced the specificity of nomadic states here should be singled out - primarily separate habitation of nomads and sedentary population in different ecological zones, and relatively high density of the nomadic population which facilitated its unification. The major political characteristic of the region was that the outside sedentary world frequently opposed the nomadic one in the form of large states, sometimes even of empires. There were also, of course, small polities and states. For example, the subjugation of isolated oases in East Turkestan, in those cases in which China was in no position to oppose this, did not require great strength. However, considerably more often, before the dependence of sedentary societies could be established and, even more so, before they could be conquered, the nomadic society itself or even several such societies had to be consolidated. This, correspondingly, entailed an increase in social differentiation in these societies. In individual cases nomadic states emerged

Nomads

and the state

on the eve of their conquest of agriculturalists and townsmen, although the future of such states depended on the success or failure of the conquest. Finally, the major economic characteristic, which is partially linked to the characteristics already mentioned, consisted in the specific difficulties involved in maintaining regular and uninterrupted trade between nomads and the sedentary population (for more detail on this point see pp. 205ff.). For the reasons which have been suggested, the subjugation of sedentary societies was very important to the nomads of the Eurasian steppes and usually took place in the form of the conquest of sedentary societies. I am going to break with chronological sequence and begin with the Mongols. This is because we have far more detailed information about the states they created than we have about other nomadic states in this region in ancient times and in the Middle Ages. Recently certain scholars have begun to assert that the state of Jenghiz Khan was not the first Mongol state and that already in the twelfth century amongst the Naimans, Kereits and even amongst Mongols proper there was a state and a class society (Istoriia Mongolskoi Narodnoi Respubliki, 1967:109; Gongor, 1973:121; Kychanov, 1974:169; Sandag, 1977; Krader, 1978:99; Krader, 1979:227). The inconclusiveness of this view is well illustrated by Munkuev (1977). Right up to the beginning of the thirteenth century amongst the Mongols, only separate tribes and tribal polities were known - the Naimans, Kereits, Merkits, Tatars and others - which essentially were no more than dispositional chiefdoms. The composition of all of them, however, was fairly fluid. 'The tribe [irgen] was a changing quantity which was very little organized and disparate. Some sort of polity would temporarily emerge in the time of wars when they were preparing to invade somebody or to repulse the invasion of a rival tribe' (Vladimirtsov, 1934:79). Correspondingly, the khans in this period '. . . were ephemeral leaders of indefinite groups with indefinite, always disputable power' (Vladimirtsov, 1934:80). At one time, in the second half of the twelfth century, there existed a polity of Mongols proper (Secret History, §52, 53, 57, 58; Rashid al-Din, 1952a:35-46). But it did not include all the Mongol subdivisions and soon disintegrated. After the death of Khutula Khan the Mongols did without common leaders, and when Temujin fought for the unification of the Mongols the Tayjiuts and other purely Mongol subdivisions joined the ranks of his enemies. For Mongolia in this period this was nothing unusual. The twelfth century was a time of fierce struggle not only between different tribal polities, but also within them, between different tribes and even their subdivisions. The Secret History and Rashid al-Din describe this in detail and in full accordance with one another. The only difference is that the Secret History 234

The Eurasian steppes, semi-deserts

and deserts

(§254, quoted in Kozin, 1941:184) adopts an epic tone to describe the intestine wars: The starry sky used To turn away For there was such discord Amongst all the people. Nobody lay in bed there. Mother earth Trembled For there was such discord Amongst all the people. Nobody wrapped up in a blanket, Everything was pierced through with swords. Each man encroached on another, Nobody lived by his own free will. Who will get out in this general scramble, Who will survive this deadly battle?

And the prosaic Rashid al-Din (1952:104, 105) relates the facts in plain language. It is hardly surprising that Temujin needed more time and force to unify the Mongol tribes than he needed in all his ensuing conquests. As usual the struggle was about pasture, livestock, booty and the submission of one group of nomads by another. It is my impression that internal struggles within twelfth-century Mongol society went further than the usual internecine wars waged amongst nomads. There are foundations for the suggestion that the balance between productivity of pastures, number of livestock and population-size was upset amongst the Mongols. At the same time as something akin to overpopulation can be observed amongst them, the dry climate of the twelfth century led to a worsening in grazing conditions (Khazanov, 1979). It is no coincidence that hunting, and sometimes fishing even, played an important role amongst Mongols of this period. No less indicative was the 1147 agreement, unfavourable for the Ch'ing, according to which the Ch'ing state had to supply the Mongols not only with agricultural products, but also with sheep and cows (Martin, 1950:58; Vorobev, 1975:330). But supplementary sources of subsistence outside the nomadic society proper were extremely limited amongst Mongols in the twelfth century. Transcontinental trade declined (Vorobev, 1975:338-9), and the relations established with China were not to the advantage of the Mongols (Martin, 1950:57-9). It is clear that in the twelfth century Mongol society was in a 'state of stress'. At the same time there was social differentiation in the society. However, there is no data which verifies the opinion of Krader (1978:99-100) that it was already divided into two classes which were differentiated by their relationship to the means of production. Besides ordinary nomads, arats,

Nomads

and the state

amongst the Mongols there was an aristocratic stratum, and there were also dependent groups (see pp. 177-8). But ordinary free nomads paid no regular and fixed taxes and fulfilled no obligatory duties for the aristocracy. Nevertheless the latter was turning into a hereditary estate. Temiijin was the great-grandson of Khabul Khan, one of the most well-known and influential chiefs of the first half of the twelfth century. However, the social positions of the aristocracy were not yet finally stabilized. The genealogy of Temiijin himself is witness of this fact. His great-grandfather headed the chiefdom which included the greater part of Mongol subdivisions proper. But subordinate to his father Yesugey, who was not yet Khan, but only Baatur, was his own subdivision and just one section of another subdivision of Tayjiuts. Temiijin himself, despite his noble descent, began his career a poor and uninfluential man without dependents or nökur. Such a situation was not unknown amongst the Mongols. Only not everyone was able to master his fate as Temiijin did. However, it was because his activity coincided with an objective tendency in the develop­ ment of Mongol society that he was able to do this. Mongol society very much needed to be united so that with the aid of conquests it could deal with its internal economic and social difficulties. So if it had not been Temiijin someone else would have completed the unification. The question is how and in what form? The year 1206, in which Temiijin became Jenghiz Khan, can be taken as the year of the birth of the Mongol state. Its future then depended entirely on successful external expansion. Without the latter the Mongol state would have been fated to disintegrate quickly. Thus it is no coincidence that even before Temiijin was proclaimed Khan, and immediately after this event he conducted his first campaigns against the Tangut (Kychanov, 1968:298f.; Kychanov, 1977). The single Mongol Empire was a nomadic state of the first type. Not by any means always, nor completely did the Mongol nomads become involved in the life of the countries they had conquered. Local people organized the collection of taxes and management of many sedentary territories. The most famous amongst these were Mahmud Yalavach and his son Masud Beg who governed Maveraunnahr. Broad external expansion still continued in this period and the pillage of conquered territories, which was ineptly controlled by the Mongol government, for the time being more than adequately fulfilled its aims (Schurmann, 1956:304). Social relations in the conquered countries underwent a specific 1

1

According to Lattimore (1963:6), after the unification of Mongolia, Jenghiz. Khan first conquered nomadic tribes and only after this did he turn to agriculturalists. The chronology ol events opposes this view: 1205-27 - the conquests of Hsi Hsia (the Tangut); 1211-34 the conquest of the Ch'ing state; 1218 - the establishment of Mongol supremacy in East Turkestan; 1219-24 - the conquest of Middle Asia, etc. The conquest of nomads occurred incidentally, as the basic blows were directed against sedentary states.

236

The Eurasian steppes, semi-deserts

and deserts

transformation, which was more marked in some cases than it was in others. However, this transformation caused no changes of principle to take place. The exploitation of the dependent classes in the conquered countries increased considerably, although in many of them old forms of exploitation were preserved or gradually reintroduced. Neither were the old ruling classes completely destroyed, although they did have to accept the dominant role of the new Mongol ruling class and forego a considerable proportion of their rights and dues in favour of that class (Schurmann, 1956). The socioeconomic relations in Mongol society itself underwent con­ siderably more changes. Ordinary nomads were forbidden free passage from one subdivision to another. Certain scholars see the establishing of feudalism, and of serfdom even, in this reform (see, for example, Vladimirtsov, 1934:110-12; Fedorov-Davydov, 1973:45; Ishzhamts, 1974:139). However, the prohibition on free passage and separation from the prescribed subdivision was also extended to the aristocracy (see, for example, the data of William of Rubruck and Piano Carpini in The Mongol Mission, 1955:27-8, 94). Fluidity in the composition of society inevitably would have put into jeopardy its military organization which, in part, was still based on former segments of sociopolitical organization, particularly the lower segments (on this point see Rashid al-Din, 1952a: 179, 270ff.). Superficial similarity to European feudalism in the organization of power should not overshadow the qualitative differences with regard to the ownership of key resources, the military system and relations between different strata and estates of Mongol society and within these strata themselves. Indeed, even with regard to organization of power the differences were very great. The Mongols knew nothing of real European vassalage. They were all vassals of the Qa'an (Vladimirtsov, 1934:105-6, cf. on the other hand 104; Markov, 1976:83; Krader, 1979:228-9). It seems to me that there were two reasons for this particular reform. The politics of conquest demanded centralization and discipline. Consequently, the Mongols were transformed from being an armed people into a people of warriors with a defined military organization based on a decimal system. Such a military organization had existed amongst the Inner Asian nomads since ancient times; Jenghiz Khan merely developed it further. Many scholars have already written on this topic (see, for example, Kychanov, 1973:81). The second reason is linked to the circumstances around Temiijin's seizure of power. Some of them have not yet been examined in sufficient detail. In his time Barthold (1963b: 137) put forward the view that Temiijin's main rival, Jamuqa, was the head of the 'democratic party' of Mongols, while Tcmujin headed the 'aristocratic party'. Available sources offer no foundation for this suggestion. Nevertheless, what is true is that the .'37

Nomads

and the state

differences between Temiijin and Jamuqa were not a question of simple rivalry, for both wanted to unite the Mongols, but on a different basis. Jamuqa wanted to preserve the traditional forms of sociopolitical organization, including different nomadic polities. Temiijin broke them up ruthlessly. It is no coincidence, therefore, that the majority of Mongol subdivisions proper for a long time supported Jamuqa, and the nbktir and army of Temiijin consisted of individuals from the most diverse subdivi­ sions, united not by traditional social ties, but by personal loyalty to the chief or simply by force. It should not be forgotten that the greater part of the traditional Mongol aristocracy, particularly the higher aristocracy, was wiped out by Temiijin during a long and bloody struggle. In some respects this struggle is reminiscent of civil wars, those very wars in which, according to their medieval expert Farinata delli Uberti, 'Faith, I was not alone there nor had gone In with the rest without good cause' . . .

(Dante, Inferno, Canto X, lines 89-90). Essentially, as this war died out a new aristocracy emerged, one which consisted of kinsmen of Temiijin, individuals who had risen from the ranks during his struggle to unite the Mongols, and only partly of the old nomadic elite. This state of affairs facilitated Temiijin's reform which was directed to the further unification of Mongol tribes. It is not known to what extent the experience of the previous nomadic states was taken into account in the steppes, or to what extent the chronic weaknesses inherent in the actual circumstances in which these states emerged were recognized. As has already been pointed out, the creation in the steppe of a large nomadic state demanded that several political units joined together. This was usually brought about through the use of force. Therefore, apart from conflicts between nomads en masse and the conquered sedentary population, there were other conflicts between privileged units of nomads and those who were pressured into submission to them, either by force or because of circumstances, but who preserved their own autonomy. Successful external expansion could pacify the latter for a time, but did not completely eliminate the conflicts. The example of all the Turkic qaghanats (see below pp. 255ff.) in this respect is very indicative. It is hardly likely that Temiijin knew anything specific about these qaghanats. The nomads of Inner Asia were more concerned with cyclical, rather than linear time. In such conditions an oral tradition adopts a legendary tone. But the experience of the internecine Mongol wars was sufficiently instructive for whoever cared to learn something from it, The unification of the Mongol tribes was achieved by force and there is no doubt 238

The Eurasian steppes, semi-deserts

and deserts

that Temujin, even before he became Jenghiz Khan, was thinking of ways of consolidating Mongol unity. Vladimirtsov (1934:96) is not completely correct when he writes that'. . . usually Jenghiz Khan introduced nothing new'. Krader (1968:95) also underestimates the differences between the Mongol state and its predeces­ sors when he emphasizes its tribal basis. The military-political system created by Jenghiz Khan was in actual fact an innovation, for it broke up the upper segments of the traditional social and political organization of Mongol nomads and partially reshuffled their different subdivisions, thus eliminat­ ing the threat of separatism, at least in the traditional form in which this occurred amongst nomads. Its effectiveness was proved in practice, but the system had nothing to do with the emergence of feudalism. Perhaps another innovation was the notion of Heaven-sanctioned supreme power of divine origin with which, according to de Rachewiltz (1973), Mongol nomads were unfamiliar and which they borrowed, as the ancient Turks had done before them, from sedentary societies. The main obligation of the Mongol arats was military service. Under Jenghiz Khan they received no reward for this service. There also were other obligations (Schurmann, 1956:312ff.; Munkuev, 1965a:70f.). Judging from the Secret History (§278-80), under Ogodey they were relaxed and regulated. Apart from postal dues, the arats also paid dues in kind (shulen) - one two year old ram from a herd, milch mares set aside for the Khan's court, and one sheep out of every hundred was given up for the poor. This last obligation, qubchur, goes back to the old reciprocal traditions of Mongol society. Rashid al-Din (1946:281; see also 1952:135) wrote: 'Earlier, when their traditions and rules still existed, all the Mongol war­ riors every year gave to the impoverished hordes and bodyguards qubchur of horses, sheep, oxen, thick felt, krufi and other things.' By now it had acquired a redistributive form and later turned into a tax levied not only on nomads, but also on subjugated sedentaries. Further­ more, a portion of war booty was given to the Khan and other members of the aristocracy, even in cases where they had not participated in the campaign (Meng-ta pei-lu, folios 12a-b, quoted in Munkuev, 1975:67-8; cf. Vladimirtsov, 1934:113). However, in 1233 an amendment was introduced that one head from the overall number of cows, sheep and horses would be taken even in those cases in which there were less than 100 head of such stock in one household (Munkuev, 1965b: 137-8). It is possible that Ghazan Khan deliberately exaggerates the difficult 2

' This concept may also he observed amongst the Scythians (Khazanov, 1975) and, consequently, may be regarded as u characteristic ideological notion in the nomadic states of the I'.unisian sleppc-s. •' A compressed, salty cheese.

2!W

Nomads

and the state

material existence of ordinary nomads under his predecessors in order to emphasize his own concern about them. Nevertheless, his words are remarkable. 'Now from no one is it hidden that in the time of our brave fathers on the Mongol ulus there were imposed different obligations and burdens, such as qubchur of stock, maintenance of large yams, the carrying of the burden of a severe yasak and qalan, which in one move we have abolished. For the most part they [the Mongols] were bereft of [food] stores and taghar; however, despite these burdens, they laboured honestly and served [the sovereign], bore the adversities of distant campaigns and were satisfied with little. There is no doubt that until the present time the Mongol army has achieved little wealth and property' (Rashid al-Din, 1957:292). At the same time it must be pointed out that, despite the view of Vladimirtsov (1934:110ff.) and his followers which include Krader (1978:102; 1979:230), there is no convincing data which shows that ownership of pastures, let alone of livestock, rested exclusively in the hands of the Mongol aristocracy. All that the latter possessed to a considerably greater extent than previously was rights to the regulation of pastoral migrations and the distribution of pastures. Towards the beginning of the thirteenth century there emerged in Mongolia not nomadic feudalism which, according to the view of Vladimirtsov, differed little from feudalism in Western Europe, but a society of another type. In this society the main differences between the different strata and classes consisted not in their relation to key resources, but in their relation to power and government. Those direct obligations which the rulers imposed on the ruled were not the cause, but the consequence of the emergence of the rulers. However, the socioeconomic situation in Mongol society acquired a specific form because on the broad imperial level relations between the Mongol aristocracy and the arats were not fundamental. The main differences, the principal forms of dependence and exploitation, were established in the relations between the Mongol nomadic aristocracy and the conquered sedentary population, both as a whole and with its separate classes. Yet more significant is the fact that even under Jenghiz Khan and his immediate successors there began a struggle amongst various groups of the Mongol elite over the problem of how to manage and exploit the conquered sedentary peoples. In one form or another this struggle continued over a long period and after the disintegration of the single Empire (Barthold, 1963:525ff.; Iakubovsky, 1946:48-52; Tolstov, 1948:344; Petrushevsky, 1952:12-15; Petrushevsky, 1960:32, 48-53; Petrushevsky, 1977:239-40; Stroeva, 1958; Munkuev, 1965:18, 44ff.; Khazanov, 1975:63). The struggle was not only about whether the exploitation of conquered countries should take on predatory contributional or tribute forms, or a 240

The Eurasian steppes, semi-deserts

and deserts

more regulated form. Whether consciously or unconsciously so, it was also about something else, namely about whether the nomadic aristocracy should establish closer relations with the ruling classes of conquered countries, or should remain the hostile conqueror relying only on military force. Conflicts over the question of relations to sedentaries had already begun amongst the Mongol aristocracy during the conquest of Middle Asia. Juchi tried to avert unnecessary destruction and the annihilation of sedentaries, but Chaghatay was merciless (Barthold, 1963:525). The 'Great Yasa' of Jenghiz Khan forbade Mongols to adopt a sedentary way of life and to live in towns. Nevertheless, the inexorable objective needs of the government of the Empire led to the foundation of its capital, the city of Qaraqorum (Kiselev and Merpert, 1965). Also significant is the fact that the immediate successors of Jenghiz Khan attempted to create a handicraft sector of the economy in their state, but again this was to the cost of the conquered. When towns were seized their craftsmen and artisans were given over to members of the ruling clan, who had a free hand over them. On some of these craftsmen the Mongols imposed an almost slave-like dependence and then forced them to move to Mongolia. In Qaraqorum, for example, captured craftsmen from Middle Asia, Iran, China and even France were to be found (Olshki, 1946). On other craftsmen the Mongols imposed a natural tax which had to be paid on the spot. Chaghatay faithfully followed the testimony of his father. On the other hand, Ogodey, according to the testimonies of Juvaini, al-Suyuti and other historians, tried to put an end to unnecessary destruction. It was he who issued the order for the rebuilding of Herat. Far more important is the fact that he understood the aphorism of the ancient Chinese orator, Lu Tsia (third to second century B . C . ) , taught him by the Chinese counsellor Yehlu Ch'uts'ai: 'Although you inherited the Chinese Empire on horseback, you cannot rule it from that position' and allowed scholar Confucians to be drawn into the civil administration. Yehlu Ch'uts'ai persuaded Ogedey to regulate the taxation system in China and in practice proved that the treasury could only benefit by doing so (Munkuev, 1965). However, the reforms involved a bitter struggle with the 'old Mongol' party. Ogodey himself sometimes hesitated and was unable to fully eliminate the arbitrary rule of the Mongol elite in conquered countries (Duman, 1977:338-9). Ogodey's heir, Guyuk, tried to return to the policies of his grandfather and uncle. According to the testimony of Juzjani, in his court serious consideration was given to a plan to completely massacre or castrate all the Muslims in the eastern countries of the Empire. But Mongke continued the political line of Ogodey. He put a limit on the taxes and dues owed by the peasants and townsmen of the conquered 241

Nomads

and the state

countries, approved a fundamental currency-reform and tried to hand over the government of the sedentary countries to the general imperial administration so as to decrease the arbitrary rule of the Mongol nomadic aristocracy (Rashid al-Din, 1960:141-2). The cessation of conquests coincided (and this was hardly due to mere chance) with the beginnings of the disintegration of the Empire. As the Empire disintegrated into separate states, the development of each of these states acquired its own individual features. However, in all of them a predominance of one of the first two tendencies may be observed, although they were put into practice with different consequences and with different degrees of success. The first type of nomadic statehood was preserved longest and in its most pure form in the Golden Horde, where there was a relatively clear geographical demarcation between the basic mass of nomads and the sedentary population. It was only immediately after the Mongols conquered the Russian princedoms that they directly interfered in the government of these, physically destroyed the most powerful and independent princes, conducted a general census of the population, established a taxation system and sent special officials - basqaqs and darugas - with military detachments into Russian towns. Later on, however, all the khans of the Golden Horde sought was to receive their regular tributes (although there were other payments and extraordinary collections), to maintain their supreme sovereignty over the Russian princedoms and to stave off dangerous harassment from any one of the latter. From the fourteenth century the Russian princes themselves brought their tribute to the Horde; they went with bribes and ready to vilify one another, indulging in different intrigues against each other. As the Princedom of Muscovy grew more powerful it sought to stand as the sole representative of all the other princedoms before the khan of the Golden Horde. At the same time journeys of the princes to the Horde became more and more rare (Nasonov, 1940; Grekov and Iakubovsky, 1950; Safargaliev, 1960; Spuler 1965; Grekov, 1975; Poluboiarinova, 1978). This policy, or variations of such, was followed with regard to other sedentary countries which had been forced to acknowledge vassal dependence (Konstantinu, 1974:163). In contrast to the nomadic aristocracy in Iran, that of the Golden Horde did not have iqta, but it did have administrative needs, and the benefits it derived from the creation of manufacturing and agricultural sectors of the economy were evident. The policy of the khans of the Golden Horde was quickly to build towns in the territories of their state which were in­ habited by nomads and to turn these towns into trading centres. In order to do this they had to introduce people from the countries they had conquered into these towns. Indeed, it was these migrants who created the 242

The Eurasian steppes, semi-deserts

and deserts

lively, syncretic culture of the towns of the Golden Horde of which, together with the town-like settlements, there were over 100. About 20 of them were large centres and had mintage rights (Egorov, 1973:12; see also Povolzhe v srednie veka, 1970). There is less information about agriculture in the Golden Horde. Apart from unfavourable natural conditions, evidently this is linked to the fact that the Mongols received the greater part of the agricultural products they needed in the form of tribute from the lands they had subjugated. In fact, Piano Carpini (The Mongol Mission, 1955:42-3) noted that large agricul­ tural economies had been created and that they were worked by individuals from the dependent population, perhaps primarily by prisoners of war who had been turned into slaves. However, according to the testimony of al-'Umari (quoted inTizenghausen, 1884:229,230,233), in comparison to Iran, there were few towns and villages in the Golden Horde. Judging from information supplied by al-'Umari (Tizenghausen, 1884:230), Barbara (Travels to Tana and Persia ,1873-21) and certain other sources, the sedentarization process began in the Golden Horde, but did not develop to any great extent there. Only a certain number of impoverished nomads and part of the nomadic aristocracy which was linked with the administrative apparatus settled in towns (Fedorov-Davydov, 1973:85-6). There developed in the Golden Horde not so much sedentary life as towns, and these are not quite the same things. Moving on now to a general evaluation of the Golden Horde, the nature of relations between the nomadic aristocracy, the ruling class in the state, and the ordinary nomads in it must be examined. Even Berezin (1864:468ff.) in his work which to this day has not lost its original value, noted the essential difference between the situation of subjugated sedentary population and that of ordinary nomads. Nevertheless, even the latter paid taxes and fulfilled other dues. According to the same al-'Umari (quoted in Tizenghausen, 1884:235), in years of natural disasters they sometimes had to sell their own children into slavery in order to pay their taxes. However, there is no data to testify to the fact that the taxes differed from that of Hulagus Iran or that they changed in any essential way in comparison with an epoch of the single Empire (Fedorov-Davydov, 1973:39-41). It is curious that with all of this the ethnic origins of the nomadic aristocracy and ordinary nomads in the Horde were different. In Eastern Europe the Mongols physically destroyed the Polovtsian aristocracy and themselves took its place (Fedorov-Davydov, 1966:228ff.; Fedorov-Davydov, 1973:27-8, 41). Nevertheless, the nature of their relations with the nomads they had conquered was in principle similar to those which existed in other states of the Jenghizids. Evidently this is linked to the general laws of the functioning of nomadic states of the first type, '('"'est le ton qui fait la chanson.' On the whole, despite those scholars who regard the Golden Horde as a 243

Nomads

and the state

feudal state (see, for example, Fedorov-Davydov, 1973), I am inclined to regard it as an early state. In the Golden Horde tribute was the principle way through which the conquered population was exploited. Slaves, particularly in the towns, formed a stratum in the dependent population, but little is known about the role of this slave stratum. On the other hand, it is noteworthy that during the course of the entire history of the Golden Horde, apart from certain temporary fluctuations, the export of slaves was of considerable importance (Safargaliev, 1960:91; Khazanov, 1976:264-6). The nomadic aristocracy and ruling classes of the vassal states remained completely isolated: it was only by force that they had been united into one political super-system. Admittedly, it has been observed that in the fourteenth century relations between the nomadic aristocracy and the newly formed sedentary elite of the new towns, particularly the trading-merchant elite, became somewhat closer (Fedorov-Davydov, 1973:82-3), but this rapprochement was incomplete and temporary. When in the fourteenth century the Golden Horde gradually began to disintegrate into separate political units, in each of the latter there were sedentary areas adjacent to the areas where nomads lived. But these areas gradually fell into decay. Consequently, in the middle of the fourteenth century the nomadic aristocracy turned with renewed vigour to seizing sedentary lands (Mordovia, the Bulghar territories around the River Kama), that is to say new external sources of exploitation, and they even tried directly to subjugate the Russian princedoms in order to become the ruling class within them. The states which arose from the wreckage of the Golden Horde developed in different ways; but amongst these states were all the above-cited variants of changes which manifest themselves after the disintegration of nomadic states of the first type appear. The Great Horde strove to continue the former policy of the Golden Horde, but this was impossible because of the changed balance of forces. Thus its disintegration and destruction were not long in coming. In the sixteenth century, the Dasht-i-Qipchak 'returns to its nomadic essence; nomadic forces predomin­ ate without restraint, having swamped all those elements of urban life which were present a hundred years previously, but which had left almost no inheritance in the steppe' (Fedorov-Davydov, 1973:166). The Astrakhan Khanate was a smaller replica of the Golden Horde. The process of sedentarization was predominant in the Crimean Khanate. Only in peripheral areas of this Khanate was any semi-nomadic economy preserved, and this was partially maintained by booty from raids on Russia, Poland and Lithuania and the subsequent sale of the prisoners on the international market. The Qazan Khanate, in which nomadism was already unfeasible on account of ecological conditions within the Khanate, was a typical sedentary state, the ruling class of which was made up of the 244

The Eurasian steppes, semi-deserts

and deserts

descendants of the former conquerors. Only certain traditions of the Golden Horde period were preserved in the political life of the state. The Mongol state in China was different in character. Already in 1230 certain members of the nomadic elite were suggesting that the population of North China should be annihilated and the country turned into pastureland (Munkuev, 1965:190). Ogodey did not go along with this and forbade the seizure of any land for pasture (Schurmann, 1956:29); Qubilay sought to avoid any unnecessary sacrifices during his conquest of South China and he ordered that no peaceable inhabitants of the country be killed and forbade arbitrary pillage (Schurmann, 1956a:306; Svistunova, 1977:290-1; Duman, 1977:341; Borovkova, 1977:447-8). Having transferred his capital from Qaraqorum to Peiping, then in 1267 to Dadu (Khan-Baliq), where present day Peking is situated, and having founded the Yuan Dynasty, Qubilay had made an important step towards turning his state into a nomadic state of the second type. The price of this was a protracted civil war with Ariq-Boge and Qaydu in which the fate of the entire Mongol state was at stake. The opponents of Qubilay sought to keep Mongolia as the centre of the Empire and themselves become khans like the former ones (Dalai, 1977:327). But Qubilay wanted to be emperor of China, even if this meant the disintegration of the Mongol empire. After his victory, even from the geographical point of view the unity of Mongol society was destroyed. The gulf between the different groups of the Mongol aristocracy, the interests of which were now diverse, became deeper. Those who had remained in the steppe were not pleased that the lion's share of the revenue from the exploitation of conquered China was concentrated in China with those Jenghizids and their followers who had discounted his testament. Herein lies the reason behind the proclivity of the 'old Mongols' for organizing uprisings and discord (Bessac, 1965:384). In its own way the Yuan state violated the laws of ecology because to the south of the Great Wall of China lay an ancient zone of irrigation-agriculture and the great mass of nomads continued to live to the north of the Wall. The only exceptions to this were those who had been called up into the army and those who had moved into China together with the dynasty. According to the Mongol chronicle of the seventeenth century, at the end of the Yuan Dynasty there were more than 400 thousand of them (Danzan, 1973:252; cf. Munkuev, 1977:400-1), that is far fewer than there were in Mongolia, but a drop in the ocean in comparison with the Chinese. Nevertheless, the Mongols in China preferred direct government to tribute forms of dependence. 'Conquest is important, but preservation no less so.' These words of Ovid, which a certain King of Aragon brought to his State Council, convinced that he was quoting the Bible, were felt by the Mongols back in the time of Jenghiz Khan. Even then it was clear to them that they could not do without .'45

Nomads

and the state

the Chinese state machinery (de Rachewiltz, 1968). Those who migrated to China understood this fact completely. Consequently they preserved the state machinery with few changes and with the Chinese bureaucrats in place, although they realized the dangers involved in relying on the latter and had no wish to merge with the old Chinese ruling classes. This explains their original aim to introduce non-Chinese ethnic elements into the administra­ tive apparatus, sometimes in the form of individuals such as Marco Polo who were equally exotic to the Chinese and the Mongols, and to limit the use of written Chinese in the affairs of the state - an attempt which turned out to be ineffective. Finally, the population of the Yuan state was divided up into four categories, each of which had different rights and privileges: the Mongols; migrants from countries west of China; the inhabitants of North China; and the inhabitants of South China. The incumbents of the central organs were usually Mongols or migrants from Inner and Middle Asia, but these leaders had to take on Chinese as their advisers. Local organs of government consisted of both Mongols and Chinese (Duman, 1977:335-6). Furthermore, the old Confucian examination-system was revived, which is witness to the fact that fundamental changes in the sociopolitical system in China were impossible. The socioeconomic relations established in Yuan China were no less contradictory than the political ones. Social differences were not always as the ethnic model proscribed them. Already under Ogodey in 1236, despite the opposition of Yehlu Ch'uts'ai, the distribution of Chinese lands as appanages to the relatives of the Khan and to important members of the military elite had begun. The appanages (shih-yi, literally 'territories for feeding' - Duman, 1977:344ff., n.71, 74) involved rights of the collection of taxes from the conferred territory and frequently they became hereditary. This Far Eastern analogy to the Near and Middle Eastern iqta sometimes consisted of whole areas and districts (Schurmann, 1956:67; Munkuev, 1965:47-8), and in China it also was a source of separatism. However, in Yuan China the appanage system played a smaller role than it did in Hiilagiis Iran. Moreover Qubilay was able to some extent to limit the power and influence of their owners (Schurmann, 1956:8; Melikhov, 1977:77-8). Large-scale private land ownership had been preserved in China from the former dynasty. But the composition of the landowning class under­ went essential changes, particularly in North China where many Mongols and their Khitan (K'i-tan), Jiirchen and Middle Asian allies now settled. However, in South China large-scale Chinese land ownership predominated (Schurmann, 1956:23; Duman, 1977:358; Borovkova, 1977:448). The economic position of landowners and even of peasants in South China was better than it was in North China. On the other hand they were not allowed into the central apparatus of power (Borovkova, 1977:450). 246

The Eurasian steppes, semi-deserts

and deserts

Little is known about the position of the ordinary nomads who remained in Mongolia and the adjacent areas (on the settlement of nomads in the latter see Rashid al-Din, 1952a:279;Vladimirtsov, 1934:125; Melikhov, 1977:73) and tried to carry on with their former way of life. Even if their social position underwent no essential change in comparison with the period of the single Empire, their economic situation worsened. 'In so far as it is possible to judge from the meagre sources which we have to hand, during the Yuan Dynasty the prosperity of Mongolia and the Mongols went into sharp decline, particularly in comparison to the time of Jenghiz Khan and his three successors. Ceaseless feudal warfare and the maintenance of the large contingents of warriors which were needed to protect the empire had exhausted the country. And meanwhile, the growth of trade capital in the hands of the "Muslim" and Chinese merchants did not create and could not create new forms of production. The newly founded towns in Mongolia, evidently, did not flourish, nor did agriculture develop . . .* (Vladimirtsov, 1934:127). Probably it was not so much the internecine wars, to which the Mongols had long been accustomed, and the economic obligations, which were scarcely more weighty than they had been under Ogedey, that affected ordinary nomads as the burdens of military and garrison duty (Dalai, 1977:336); these duties involved them in a great deal of expense and prevented them from carrying on their customary economy in the optimal habitat. According to Marco Polo (Ch. LXXVIII, Ramuzio ed.), already under Qubilay the wages paid to the warriors did not cover their expenses. Frequently they lived on the proceeds of their own herds and the milk they sold to townspeople. All this led to the situation in which wives and children sometimes had to be sold into slavery. Thus, in 1317 according to the court record, the Emperor ordered: 'Recently it has been said that the Mongol tribes have become impoverished and often sell their sons and daughters . . . into slavery. Thus we order the officials to buy them up and return them to their tribes' (Yuan Shi, Ch. 26, quoted in Munkuev, 1977a:423). The government recognized the danger inherent in this kind of situation and tried to ease the material burden on ordinary nomads, but without any particular success (Munkuev, 1977a). Even worse was the situation of those Mongols who had migrated to North China where there was insufficient pasture for them. Many of them 4

5

1

This is confirmed by contemporary archaeological research (Drevnemongolskie goroda, 1%5). On the territory of the Mongolian Peoples' Republic 23 sites which were inhabited in the thirteenth and fourteenth centuries are now known; two more fortified sites have been discovered in the Trans-Baikal area and two unfortified settlements in Tuva. The overwhelming majority of them had not even survived until the end of the fourteenth century. Many centuries earlier the ancient Turks who had been forced to settle in North China had also suffered from this. Bilgii Oaglian pathetically warned his compatriots: 'O Turkic people, when you go to this country you are on the brink of ruin' (Kliashtorny, 1964:24).

247

Nomads

and the state

had less than 100 sheep, but still qubchur was taken from them. Thus a series of government decrees was directed at easing their tax burden. In 1304 the minimum holding of livestock for tax liability for the Mongols was set at 30 head, and in 1312 at 100 head (Munkuev, 1965a:75ff.). But evidently none of this met with any great success. In this way, in Yuan China the tendency to preserve traditional social and economic relations was revived. The main changes occurred in the composition of the ruling classes which became ethnically diversified. In the conquered country the nomadic aristocracy had to share part of its power and privileges with its sedentary predecessors, but neither did it want, nor was it able to merge with those predecessors into a single class. In its capacity as the Mongol aristocracy it sought the support of ordinary nomads. But since at the same time and in the first instance it was the ruling class of a sedentary state it was compelled, whether or not by choice, to treat those nomads as one of the dependent classes. The Yuan state was a state of the second type and certainly not an early, but a developed state in which the social relations of the single Empire period survived, if at all, only on the social and geographical peripheries. But through its origins it was linked to the traditions of a conquering state, and the stamp of its origins was never completely erased. The policies of the dyn­ asty were undecided and the ensuing situation turned into a knot of political, social and ethnic conflicts, both between different classes and estates and within them. No one gained from this situation. A reflection of the social crisis was the crisis within the Mongol ruling class itself, the struggle for power be­ tween the different groups and the disintegration of imperial power (Dalai, 1977:332-3). The fall of the Yuan dynasty became inevitable. The fortunes of Mongol statehood in Turkestan were also complex. While the Chagatay ulus was separating from the empire some of the princes opposed to the uncontrolled pillage of the local population, others in fact encouraged it. Thus in the sixties and seventies of the thirteenth century East Turkestan was devastated, despite the fact that Jenghiz Khan had himself forbidden any pillaging there. The Nestorian monks, Mar Yabalaha and Rabban Sauma, on their way to China in the seventies of the thirteenth century saw that Khotan and Kashgar had been ravaged (Pigulevskaia, 1958:67). At the quriltay of 1269, on the banks of the River Talas, which formalized the formation of the Mongol state in Middle Asia, adherents of moderate policies had the upper hand. The right of every Jenghizid to a share of income from the exploitation of sedentary areas was stressed. But at the same time it was decided that they should not interfere in the government of these areas, should not collect taxes except for those which had been specified, that they should live in the steppes and mountains, migrate along specific routes and not allow their herds to trespass on arable land 248

The Eurasian steppes, semi-deserts

and deserts

(Barthold, 1963a:69). Thus, there were clear-cut features of a nomadic state of the first type in this new state. In practice not all the princes followed the established policy. The devastation of towns and agricultural territories continued right up to the ascent to power of Timur, particularly during the constant internecine wars. The disturbances of the end of the thirteenth and the beginning of the fourteenth centuries completed the destruction of the agricultural territor­ ies. al-'Umari wrote in the first half of the fourteenth century: 'In Turkestan now only ruins are to be found, which have been more or less well preserved. In the distance you see a well constructed village, surrounding which everywhere there is flowering verdure. As you get closer to the village and expect to meet its inhabitants you find that the houses are quite empty. All the people living in this country are nomads and in no way engaged in agriculture' (quoted in Barthold, 1963a:74-5). Admittedly, even in the second half of the thirteenth century individual khans from time to time tried to rely on the sedentaries of Maveraunnahr, but again and again this policy was successfully opposed by the nomadic elite. The situation changed somewhat under Khan Kebek (Kopek) (1318-26) who moved, along with many members of the nomadic elite, into sedentary areas where they became involved in direct government of agricultural territories and towns. A number of ordinary nomads migrated to Maveraunnahr with them. In exchange for military service whole districts were allotted to different subdivisions of these nomads on the right of soyurghal, that is complete tax exemption (Gafurov, 1972:466-7). But others continued to lead a nomadic life and looked upon the new trends with disdain (Barthold, 1963b.T53; Barthold, 1963c:262-5; Gafurov, 1972:466-7). Recurrent discord ensued. Khan Tarmashirin (1326-34), Kebek's brother, converted to Islam and attempted to lead an urban life, but he was deposed and killed. Nevertheless, in Maveraunnahr the second tendency in the development of the nomadic state gradually became the principal one. Under the Timurids it finally became predominant. However, it was the Chaghatays, primarily the Barlas, upon whom Timur most relied for military support. Timur gave them different privileges, specifically exempting them from (axes and other dues (Makhmudov, 1966:252, 263). However, the son and .successor of Timur, Shah Rukh, wanted to be a fully sedentary Muslim ruler. 6

" The need for cultural adaptation to the outside world meant that nomads had to acknowledge und/or accept its religions. However, the acceptance by the nomads in the Eurasian steppes of one or other religion and also their choice of the requisite moment at which to do so was a complex occurrence, which sometimes depended o n the specific historical and economic situation in the region as a whole. Any generalizations about this subject from materials concerning Eurasian nomads should be made with great care and be less specific than those made from the material about the nomads of the Near and Middle East. I intend sometime in the future to devote a special work to this very subject.

;?.49

Nomads

and the state

He rejected the laws of Jenghiz Khan and forced his subjects to follow precisely all the orders of the Sharia (Barthold, 1963b:157, 161). In Middle Asia the ruling estate of nomads finally became the ruling class of the sedentary population and merged with those of its predecessors who still remained. Ordinary nomads became partly sedentary and partly formed a distinct class of the tax-paying population, preserving individual privileges only in so far as they comprised a military estate. It is noteworthy how the meaning contained in the term 'Chaghatays' gradually changed. In the thirteenth and fourteenth centuries this was the name given to the Turkic-Mongol entourage of the descendants of Chaghatay; in the fifteenth century it designated the Turkic-speaking population of Maveraunnahr and Eastern Khurasan, amongst whom the nomadic traditions were still to a considerable extent preserved. But later on during the struggle between the Shaybanids and Timurids this was already the term used for the entire population of Maveraunnahr, both Turkic and Tadjik, both sedentary and non-sedentary, which unanimously kept their distance from the new immigrants from the Dasht-i-Qipchak, the nomadic Uzbeks of Shaybani Khan (Iakubovsky, 1941:10). In China the Mongols remained Mongols, but frequently ceased to be nomads. In the Golden Horde they remained nomads, but ceased to be Mongols. In Maveraunnahr they ceased being both Mongols and nomads. Already in the middle of the fourteenth century the single Mongol state in Turkestan divided into two parts, Maveraunnahr and Moghulistan. If Maveraunnahr finally turned into a nomadic state of the second type, then in Moghulistan the transformation occurred in a different direction (see Barthold, 1963a:79-85; Pishchulina, 1970). The agricultural-urban basis of this state from the beginning was too small. It consisted only of East Turkestan, although neither continuously nor in its entirety, and Semirechye where sedentary life dragged out a miserable existence. Thus all the attempts of khans to follow the example of Maveraunnahr were invariably unsuccessful. The fate of the state of Moghulistan depended on a series of extra-political factors, primarily on its ability to expand. But even here internal conflicts were to be found. Khans strove towards firm subjugation of sedentary territories, the nomadic aristocracy towards the pillage of such territories. As a result the khans of Moghulistan were unable to seize sufficiently large sedentary territories and even less were they able to create an agricultural-urban centre within the society itself. The most influential force in Moghulistan was the nomadic aristocracy, begs (emirs in Persian literature), who headed the different nomadic units. To ordinary nomads they acted as the leading estate (but not class), and to the dependent sedentary population as the ruling class. It is doubtful, however, whether we can generally talk of the one and a half century of Moghulistan's existence in terms of the continuous existence 250

The Eurasian steppes, semi-deserts

and deserts

of one state. At times the only thing that remained of this state was the khan's power, and then this was more nominal than anything else. The khans and their supporters tried to rely on the towns and sedentary territories, but nomadic groups successfully opposed them in this. It is noteworthy that the leaders of the Dughlat tribe which owned the oases of East Turkestan frequently encouraged centralist tendencies in Moghulistan. In the second half of the fifteenth century when, after a period of continuous discord, Yunus-khan was established on the throne, according to Muhammad Haidar: 'The Moghul people and the emirs extracted from the Khan the pledge that henceforth he would not force them to live in towns and populated cultured places because it was this which was the reason for the discord amongst the Moghuls and their grievance against the Khan' (Muhammad Haidar, quoted in Materialy po istorii kazakhskikh khanstv, 1963:201). Yunus-khan finally established himself in Maveraunnahr, but he lost his power over the nomads. At this time the end of Moghulistan was fast approaching. It is curious that at the end of the fifteenth century the power of the khans of Moghulistan extended over certain sedentary areas, rather than nomadic ones. In the beginning of the sixteenth century what was essentially a new sedentary state, Moghulia, emerged, in which the dynasty was of nomadic descent; and in the first half of the sixteenth century the Moghuls' ascendancy over Semirechye finally ceased (Pishchulina, 1977; cf. Zlatkin, 1964:64-6). The Mongol state in Iran was created when the struggle about further development was already quite clearly established amongst the Mongol aristocracy. At first the Hiilagus basically adhered in their policies to the first line, that of Jenghiz Khan and Chaghatay (of the extensive literature devoted to Mongol Iran see Lambton, 1953; Spuler, 1955; Ali-Zade, 1956; Petrushevsky, 1960; Petrushevsky, 1977; Petrushevsky, 1968; Boyle, 1968). Nomads did not become sedentary if only for the fact that the situation of a warrior-nomad was incomparably better than that of the peasant, raiyyat, who had no rights, and was oppressed and exploitable. The Il-Khans themselves pursued a semi-sedentary way of life. The conquest of Iran brought with it a sharp decline in irrigation and agriculture, an end to the cultivation of certain lands and the influx of a considerable number of nomads from the Eurasian steppes. Vassal states were preserved only on the borders of the Il-Khan state and they paid heavy dues (Spuler, 1955:357-9; Galstian, 1977). In Iran proper the Mongols established direct rule. From the very 7

' II is ixil because this Middle Eastern state was created, as were many others, by Eurasian nomads thai it is examined in this section. In the political sphere, as in all the others, Middle Hasten) nomadism is to some extent marginal and intermediary. The main reason why I do this Is that I wisli to examine all the stale of the .Jenghizids in one section.

251

Nomads

and the state

beginning their state was a conglomeration of Mongol early state institutions of the imperial period, together with the considerably more developed institutions of a sedentary Muslim society, the discrepant combination of Yasa and Muslim law. Although the composition of the ruling classes underwent considerable changes many Iranians remained within it. The Mongols preserved the bureaucratic apparatus, an inheritance from the Khwarazm-Shahs' state. This apparatus consisted exclusively of Iranians. Also preserved was the ancient division of the leading classes into 'people of the sword', who were concerned with strong centralized power only as long as conquests continued, and 'people of the pen', the traditional bearers of centralist tendencies. 'People of the sword' consisted of the politically dominant nomadic elite of Mongol and Turkic descent and of those members of the provincial sedentary elite of Iranian origin who had not been killed during the conquest or after it. 'People of the pen' consisted of the clergy of all religions and the Iranian bureaucracy. As a result of the redistribution of lands, the Mongol and Turkic aristocracy acquired cultivated lands from the Il-Khans, but they exploited the peasants on these lands and did not break-away from their nomadic way of life. They simply exacted an advantageous rent from the peasants employing arbitrary and violent methods in the process. In the first period of the Hulagtis state the nomads actually considered they were a military camp in a conquered, but still alien and unfriendly country. The reforms of Ghazan Khan and his reknowned minister Rashid al-Din, politically the Iranian Yehlu Ch'uts'ai, did not lead to the sedentarization of large numbers of nomads, but it nevertheless signified a partial shift to the second tendency of development. They abolished the farming out of lands which had flourished under the first Hulagus and ruined the peasantry and they tried to instigate fixed taxes, to ease the burden on peasants, revive agriculture and trade and so on. Furthermore, these reforms increased the political weight of the Iranian bureaucracy. Ghazan Khan did not trust the nomadic elite and executed many of them, although he still had to continue to reckon with them. The words this wise and humane leader addressed to the nomadic elite are carefully chosen: 'I am not on the side of the Tadjik ra'iyyat. If there is a purpose in pillaging them all, there is no-one with more power to do this than I. Let us rob them together. But if you wish to be certain of collecting grain [taghar]* and food [ash] for your tables in the future, I must be harsh with you. You must be taught reason. If you insult the ra'iyyat, take their oxen and seed, and trample their crops into the ground, what will you do in the future? The obedient ra'iyyat must be distinguished from the ra'iyyat who are our 8

the taghar - payment in kind of the military personnel of the state out of taxes.

The Eurasian steppes, semi-deserts

and deserts

enemies. How should we not protect the obedient, allowing them to suffer distress and torment at our hands?' (Petrushevsky, 1968:494). However, it is indicative that amongst nomads there now emerged a 'burning desire' to own estates and 'to set up in agriculture' (Rashid al-Din, 1957:291). Thus they adapted to the changing conditions when conquests ceased and could no longer act as the main source of income. In 1303 at the direct request of his army, Ghazan Khan was forced against his will to issue a yarlyq about the allotment of iqta. They gave the right to land and water and to the proceeds from them and included full tax (although not yet administrative immunity), with the exception of a small quantity of grain which the owners of iqta had to place in the coffers. Iqta consisted of large allotments, frequently whole agricultural districts, which were given to the subdivisions of nomads which constituted separate military units. Further they were divided amongst all those with military obligations, including ordinary nomads. Apart from iqta, the distribution of which was linked to obligations of military service, there also developed the distribution of mulk, which in this instance entailed hereditary, uncondi­ tional ownership. The reforms of Ghazan Khan were not final. The struggle between the adherents of both tendencies continued under the last Hiilagus. However, it would seem that the view of Spuler (1955:409) that: 'The development went astray in its early stages: Ghazan's activities would seem to have left no more lasting traces behind them' is too categoric. It is indicative that after the disintegration of the Il-Khanid State, the Chobanids, acting in the spirit of the first Hiilagus, maintained their power for a very short period. The Chobanid state was the most ephemeral and was destroyed by the Golden Horde Khan Janibeg (1356) with the support of the local elite. The Chobanids were replaced by the Jalayrids who struggled to rule in the spirit of the political tradition of Ghazan Khan. In Hulagiis Iran the nomadic aristocracy was basically the ruling class of the sedentary population, however it acted only as the leading estate of the ordinary nomads there. The lands of the nomadic tribes (yurt) were looked upon as their collective property. The tribal leaders (emirs) led the pastoral migrations, but they did not have the right to alienate grazing or to withdraw from nomads the right to use grazing territories. Admittedly, the nomads paid the traditional qubchur to the state (Lambton, 1953; Petrushevsky, 1960:361; Barthold, 1966:331) and were under the obligation to provide military service. But it should be remembered that initially they received payment, albeit irregularly, and a share of war booty. When conquests came to an end all, without exception, of those nomads who had participated in them received iqta. The nomadic sub-society in Iran finally took shape under the Hiilagus and continued to exist with different variations right up to the twentieth century. 253

Nomads

and the state

Although it was partially drawn into the developed state and class life of the country, the conditions on which the nomadic sub-society did this were largely to its own advantage, and in the political sphere it was done primarily through the mediation of its own aristocracy. This sub-society was not fully integrated in the social and economic system of sedentary Iranian society and frequently was a cause of devolution in the political system of Iran. It is remarkable that all the enumerated types, tendencies and variants in the emergence and evolution of nomadic statehood in the Eurasian steppes which can be traced in the states Of Jenghizids were known back in antiquity (for more detail on this subject see Khazanov, 1981). Both the First and the Second Scythian states were early states, based on the conquest and exploitation by nomads of sedentary peoples. But already during the Second Scythian state the sedentarization of nomads began to take place. The Third Scythian state which emerged after the Scythians had experienced a series of defeats and, correspondingly, a sharp reduction in opportunities for the exploitation of the subjected population began to develop primarily as an agricultural-urban society. Internal forms of dependence here predominated over external ones. This development was facilitated by the prolonged and multi-faceted contact with the Greek poleis to the north of the Black Sea, and the lucrative export of grain to the Greek states was an additional stimulus to agriculture in the Third Scythian state (Khazanov, 1975; Khazanov, 1978). For a long period the Hsiung-nu state developed in the same direction as the Scythian. Under Shan-yii Mao-tun many nomadic tribes and agricultural territories were subjugated, and in the first half of the second century B.C. China had to pay regular tribute after humiliating negotiations which have gone down in history as 'negotiations about peace based on kinship'. Archaeological data witnesses the wealth of Hsiung-nu society in the period of its greatest power (Davydova, 1975:142). As Ssu-Ma Ch'ien writes, Laoshan, Mao-tun's son, was persuaded by a Chinese renegade to impose for the first time taxes on the Hsiung-nu {Shih Chi, Ch. 110, quoted in Taskin, 1968:45). However, this is the only reference to this fact in the sources, and the tax burden in Hsiung-nu society proper was scarcely of any essential significance. Furthermore, the Hsiung-nu began to develop agricultural and handicraft sectors in their society, partly by using Chinese prisoners and deserters (Pan Ku, Han-shu, Ch. 70, 94a, 94b - see Taskin, 1973:24, 30, 41, 57, 121, 137). In the Trans-Baikal area the site of a Hsiung-nu town, which was a large trade centre, has been excavated (Davydova, 1968; Davydova, 1977:83-4). Nevertheless, the Hsiung-nu state remained to its end an early state. The Hsiung-nu were proud of the fact that they 'create a state by going into battle on horseback' (Pan Ku, Han-shu, Ch. 94b, quoted in Taskin, 1973:34), and in contrast to the Scythians they continued to ride the horse. Results 254

The Eurasian steppes, semi-deserts

and deserts

followed accordingly. When the Hsiung-nu began to suffer defeats, their state disappeared. Those late Hsiung-nu tribes, which in the fourth and fifth centuries A . D . migrated along with other nomads to China and there created their short-lived states, admittedly did stop riding horses, but soon after this they were quickly assimilated by the Chinese (Gumilev, 1974; Kriukov, Maliavin, Sofronov, 1979:25ff.). A very typical nomadic state of the first type in the Middle Ages was the Turkic qaghanat (551-744), an empire which, when it was at its prime, stretched from Byzantium to Manchuria (on the ancient Turks see Pelliot, 1929; Bernshtam, 1946; Bichurin, 1950; Malov, 1951; Malov, 1959; Liu Mau-tsai, 1958; Kliashtorny, 1964; Gumilev, 1967; Ecsedy, 1972). The qaghanat emerged as the result of the eminence of the Turkic (tiirk, Chinese turkut) tribal polity proper and the subjugation by this polity of many other nomadic polities and sedentary territories. The main form of relations which were then established between conquerors and conquered were different types of vassal-tribute dependence. The Turks strove to incorporate the nomadic units and polities they had subjugated into their military and political organization. In runic inscrip­ tions all these units are sometimes referred to by the word tiirk (Kliashtorny, 1964:19 n . l l ; Ecsedy, 1972:247 n.3). However, in the nomadic units which were subjugated by the Turks traditional forms of social and political organization underwent few changes. The local aristocracy was preserved although it now had to be subordinate to the state's ruling Ashina clan. Vassal nomadic units paid tribute and participated in wars. Often they were put in the front line as their loss would mean less. For this reason in all the history of the Turkic state separatist movements of nomadic vassals are very much in evidence. In the conquered sedentary areas there were tuduns, deputies of the qaghan, who collected fixed dues. However, the sources also refer to 'illegal dues'. According to the sources it was the Turkic qaghans and aristocracy who in the first instance exercised the military and organizational functions. It is known that both qaghans and the aristocracy organized the redistribution of pastures, particularly in territories which had been conquered. The qaghans also conducted sacral functions. In runic inscriptions the qaghans frequently emphasize their concern about their own people, and about the fact that they distribute amongst their people tribute, stock and slaves taken from the subjugated peoples for which in return they demand 'labour and strength' (Malov, 1959:35, 40). Most of all the qaghans were afraid that individual groups of nomads would emigrate to other territories, that is that they would lose their power over nomads they controlled. Bilga Qaghan explained the demise of the First Eastern Turkic Qaghanat by the fact that the Chinese had incited 255

Nomads

and the state

people and begs to betray the qaghan, and the people to betray the begs; he then came to the edifying conclusion that: 'If you, Turkic people, do not separate yourself from your qaghan, your begs and your homeland. . .you yourself will live happily, you will live in your own houses, you will live without worries' (Malov, 1959:24). Evidently the Turkic aristocracy was a class only in so far as conquered countries and peoples were concerned. In its own el (al, il- a tribal unit or polity, frequently used to mean the Turkic state) the Turkic aristocracy was the leading estate. It is possible that its position with regard to ordinary nomads changed over the course of the history of the different ancient Turkic qaghanats. The boundary between the leading estate and the ruling class was changeable in ancient Turkic nomadic states. In many ways it depended on the external political situation and the amount of revenue derived from subjugated sedentary areas and peoples. The history of the fall of the First Eastern Turkic Qaghanat is extremely indicative. Relations with China played a very important role here. When China was weak it could be freely robbed and forced to pay tribute. This suited the subjugated nomadic units and they tolerated this situation. But the last qaghan, Heli, because of his expenditure on arduous and unending warfare, of external political misfortunes, the defection of a number of tribes and natural disasters, needed resources. Since Heli could not cover his expenses with external revenue, according to Chinese sources, he imposed high taxes on the tribes. As a result of this the life of these tribes became intolerable and they soon rose in revolt (Lui Mau-tsai, 1958, i:143). As is made clear in sayings attributed to qaghans of the Second Eastern Turkic Qaghanat, these qaghans had been able to learn something from history and the situation in their state was different from that in the First Qaghanat. Both ordinary Turks and their aristocracy remained nomads. After the destruction of the First Eastern Turkic Qaghanat they had been forced to settle in North China, and there had begun to be partially engaged in agriculture. After their anti-Chinese uprising and the setting up of the Qaghanat, they were only too pleased to return to their former way of life. But like their predecessors (the Hsiung-nu) and their distant successors (the Mongols), the Turkic aristocracy strove to create agricultural and handicraft sectors in their society. In contrast to the Mongols, however, the Turks encouraged the voluntary creation on their territory of large Sogdian colonies which engaged in agriculture, handicrafts and trade, and even founded towns (Pulleyblank, 1952; Kliashtorny, 1964:114-22). Mutual benefit was derived from the strength of the ties which were established. The qaghans, thanks to the Sogdians' experience in trading and their connections, were able to start up the sale of war booty and tribute, particularly of silk. For the Sogdians Turkic power was no great burden. The 256

The Eurasian steppes, semi-deserts

and deserts

strength of the Turks guaranteed their safe passage along trade routes, and the political influence of the Turks assisted them in opening up new markets. A situation emerged which, neither before nor after, appeared in the nomadic states of the steppes of Eurasia. However, when Qaghan Heli, under the influence of his Sogdian advisers who wanted to continue fighting China, tried to confront the nomadic Turkic aristocracy, he met with nothing but troubles. It was only in the Western Turkic Qaghanat that the second tendency of development emerged at all clearly. Admittedly, the basic mass of Turks, in contrast to their predecessors - the Kushans and Epthalites - still continued to live in the steppes. But the local leaders in Middle Asia gradually began to be replaced by Turks or to become the deputies of the qaghan. The Turks began to move into the conquered areas, some of them settling in the oases, while others engaged in pastoralism in more outlying districts (Gafurov, 1972:222-3). However, in the Western Turkic Qaghanat the power of the qaghans in the second half of the seventh century and the first half of the eighth was weak. One disturbance followed another. The dependence of certain areas, of Sogdian Semirechye, for example, was almost nominal; at least it did not exceed the paying of tribute. Consequently, the aforemen­ tioned second tendency could not be fully realized, or else it had no time to become so before the Arabs conquered Middle Asia. The history of the Inner Asian empires, from the Hsiung-nu to the Mongols and even more so to the Manchus, was nothing like a single evolutionary line in which each new state achieved a stage of development higher than that of its predecessors. In individual periods there was, of course, continuity. But it was frequently interrupted or even sometimes turned backwards. Furthermore, the development of successor states frequently took place in directions different from those of their predeces­ sors. In the sociopolitical respect, the Uighur Qaghanat (744-840) which emerged on the ruins of the Second Eastern Turkic Qaghanat was at first in many ways similar to its predecessor (on the Uighurs see Bichurin, 1950; Hamilton, 1955; Tikhonov, 1966; Pinks, 1968; Mackerras, 1968; Maliavkin, 1974). Following the old Turkic tradition the Uighurs, who themselves formed a nomadic polity, placed other polities - Qarluq, Telengut, Tatar, Khitan, Qirghiz and a number of others - in a position of vassal-tribute dependence. The sociopolitical organization of the latter in practice underwent no essential changes. According to Chinese sources, the Uighur sent special officials to their nomadic vassals to make sure that the annual tribute was collected (Malov, 1959:41; Maliavkin, 1974:35). But some of the subjugated tribes, the Basmil and Eastern Qarluq, for example, from a judicial point of view were looked upon as the equals of the Uighur proper. ."•SV

Nomads

and the state

During the uprising of An Lu-shan and the ensuing disturbed period, which almost destroyed the T'ang, the Uighur managed to bring ruin to the country. China, which became a huge hunting ground for the Uighur, was forced to endure pillage, to pay enormous contributions, coyly referred to as 'payments of expensive gifts in return for aid extended to her', and tribute from which she was released fully only after the downfall of the Qaghanat. According to the Chinese, '. . . the Uighurs were a great calamity of the T'ang epoch' (Maliavkin, 1974:43). The Uighurs already had real towns and in the river valleys they were developing agriculture. True, it is unclear who exactly worked the land. Evidently in Uighur society proper social stratification was developing because of external factors and external sources of revenue. Shao Po-wen wrote at the beginning of the Sung epoch: 'In the beginning the customs of the Uighurs were distinguished by their simplicity and honesty and the differences between the Princes and the people were not very great. Thus the intentions of the people were single and [the Uighurs] were strong and invincible. Of course, they rendered services to the T'ang state. The T'ang [court] made rich payments. Qaghan Den-li [i.e. Tengri, 759-79] was the first to have pretensions, he began to build palaces to live in. The women had powder, colour for their eyebrows and [clothes] decorated with embroidery, all this was made in China. The squandering brought about the break with barbarian customs' (Wen Jian Lu, quoted in Maliavkin, 1974:92). Internal squabbling caused by rivalry between different groups of the Uighur aristocracy, the desertion of a number of nomadic units and natural disasters led to the destruction of the Uighur Qaghanat by the Qirghiz and the enforced emigration of some of the Uighurs into East Turkestan (Maliavkin, 1972). There the Uighurs were able to subjugate and in time to assimilate the local population. The different Uighur polities which emerged at the end of the ninth century joined together to form the Uighur Turfan Princedom. But this state developed very much along its own lines. In East Turkestan opportunities for leading a pastoral nomadic way of life were relatively limited. The Uighurs began to settle in towns and to engage in agriculture (von Gabain, 1973). In these conditions internal forms of dependence for the first time began to play the decisive role. A typical early nomadic state of the first type was also the Khitan (Qara-Khitay) state in Middle Asia. According to Ibn al-Athir, after the Khitans conquered East Turkestan and Maveraunnahr they 'trampled heavily upon their peoples', but they were limited to purely tribute forms of dependence of the sedentary population and themselves continued to pasture their herds in Semirechye. Their conquest did not cause havoc or destruction but was still regarded with hostility, as Muslims had been subjugated by non-believers. The Muslims then did not know that the 258

The Eurasian steppes, semi-deserts

and deserts

Mongols would follow the Khitans; if they had known their evaluation of the latter might well have been different. According to the same source, amongst the Khitans: Tn every town [there is] a deputy who gets money for them, and they live in kibitkas according to the custom [which existed] before they assumed power' (quoted in Materialy po istorii kirgizov i Kirgizii, 1973:73). Local dynasties continued to exist everywhere and in certain states, in Khwarazm, for example, representatives of the Khitan Ghurkhan only made occasional appearances to receive their tribute. In the middle of the twelfth century the Ghurkhan ordered his deputy in Bukhara to refer to the local ruler in all matters (Barthold, 1963b:133-4). Nomadic states of the second type were represented in ancient times by the Kushans Empire, in the early Middle Ages by a number of states created by nomads in China in the fourth to the sixth centuries A . D . , and later by the Khitan and Jiirchen states in China, the Qarakhanid and Shaybani states in Middle Asia. The rulers of the short-lived states which were created by nomads in China in the fourth to the sixth centuries A . D . at first tried to preserve nomadic cultural traditions and to prevent the spread of Chinese influence. But in the second half of the fifth century they changed political course and tried to cultivate the Chinese aristocracy, preserving the privileges and status of the latter. Thus in the Northern Wei empire which was created by the Tabgach (Hsien-pi), the Emperor T'o-pa Hong at the turn of the seventies and eighties of the fifth century issued a series of edicts which aimed at the Sinification of the Tabgach. Amongst these edicts was one prohibiting Hsien-pi dress, another even forbade conversations in the Hsien-pi language. Ordinary nomads lost all their privileges and had no prospect of social mobility in a state which, as was the case before the conquests, required its bureaucrats to be learned and well-mannered. The areas which were inhabited by the Tabgach became very provincial, places to which criminals were sent and in which the indigenous population was robbed by foreign commanders sent from the capital. Admittedly, there was fierce resistance to this policy, but the ensuing 'renaissance' proceeded only for a short time (Kriukov, Maliavin, Sofronov, 1979:32-4, 91-3). Like the Uighur qaghanat, the Qarakhanid qaghanat may be regarded as the successor of the ancient Turkic qaghanats, although in the sociopolitical respect they already appear to be quite different (on the Qarakhanids see Barthold, 1963; Pritsak, 1950; Pritsak, 1954; Kliashtorny, 1972). In the sixth and seventh centuries the Qarluqs were one of the nomadic polities which were subject to the Turkic qaghans, and at the head of the Qarluqs there were members of the house of Ashina. In the Second Eastern Turkic Qaghanat they, together with the Uighurs, joined up with the Toquz Oghuz and took 259

Nomads

and the state

part in the victorious uprising which brought about the end of the Qaghanat. After this they broke away from the Uighurs and were able to take possession of Semirechye (766), Kashgaria and partly also Farghana (766-75), but they had to acknowledge their vassal dependence on the Uighur Qaghanat. Then, immediately after the fall of that Qaghanat in 840, the ruler of the Qarluq polity, which united the Yaghma, Chikils (Jikils) and Tukhsi, as the descendant of the Ashina declared himself qaghan. Within the qaghanat power rested in the delicate balance between the Chikils and Yaghma. There were even two qaghans, of which the Chikil one was considered to be the senior. In comparison with former nomadic states, here in principle there was nothing which was new. However, for reasons which require separate research, in the second half of the ninth century, in the heart of the nomadic part of Inner Asia, the Mongol steppes, there was a demographic vacuum which was only just beginning to fill up with Mongol-speaking tribes (Khitans and Tatars). The centres of the political life of Turkic-speaking nomads moved to sedentary Muslim states. The emergence of these new nomadic states had occurred in conditions which had not previously existed. The Qarluqs were no longer pure nomads, or at least not all of them were. As before they tried to subjugate sedentary areas, but now on a firmer basis than just tribute relations. An external manifestation of their new ties and aims was their acceptance of Islam in the middle of the tenth century which, as it were, underlined their desire to take the ways of the Muslim world seriously. The Qarakhanid invasion of Middle Asia brought with it neither destruction, nor the immediate mass migration of nomads. All that can be said is that individual groups gradually infiltrated the territory of Maveraunnahr. The overthrow of the Samanid dynasty by the Qarakhanids was regarded with complete indifference by the masses who followed the advice of their clergy and came to the conclusion, according to the assertion of Hilal al-Sabi, that'. . . when there is a struggle for earthly blessings then Muslims have no need to lay themselves bare for slaughter' (quoted in Barthold, 1963:329). From the very beginning this new state was a nomadic state of the second type. However, the social changes which occurred in Turkestan should not be underestimated. The ruling class was substantially increased by the military aristocracy of nomadic origin. Essentially it was divided into two parts, the military and the bureaucratic, the latter consisting of people from an agricultural-urban milieu. As a class of landowners the dihqans were not exterminated, but they lost their political weight and economic strength. Those who recently had been nomads wanted to own land and the Qarakhanid rulers confiscated the property belonging to the most influential members of the old sedentary 26(1

The Eurasian steppes, semi-deserts

and

deserts

elite (Gafurov, 1972:403). It would appear that iqta, which had existed under the Samanids, became more widespread in Middle Asia under the Qarakhanids. This system now developed on the fertile soil of the old appanage traditions of ancient Turkic nomadic statehood, and it turned out to be very useful to the nomadic elite as a means for controlling the Iranian-speaking peasantry. However, the social conflicts in the Qarakhanid state by no means fully coincided with the ethnic ones. Despite the fact that many khans led a semi-nomadic life (Barthold 1963:378) they were closely linked to the apparatus of government, saw to the flourishing of towns and made sure that nomads did nothing to offend agriculturalists. The nomadic begs favoured separatism, for they were not content with the khans' policy towards townsmen and they united with the urban clergy against the khans (Belenitsky, Bentovich, Bolshakov, 1973:348-9). It is hardly surprising that in these circumstances the gulf between the dynasty and ordinary nomads widened and that the latter should sometimes voice their dissatisfaction. (Barthold 1963:380). But it was far from the case that in the Qarakhanid state nomads made up a serried front. There were no united tribal polities such as those which had played so important a role in the ancient Turkic qaghanat. According to Mahmud Kashgari, nomads were divided up into separate far-flung groups, some of which had become sedentary or were in the process of becoming so. For example, some of the Chikils continued to lead a nomadic life while others lived in the town of Taraz. In the middle of the eleventh century the opposition between Iran and Turan, that is between sedentarism and nomadism, the ethnic foundations of which are different, lost its meaning in 'the house of Afrasiyab' (meaning Middle Asia). Barth (1973:19) has noted: 'States have a "nomad policy", whereas nomads, since the days of the Mongols, can hardly be said to have an "agrarian policy".' This is scarcely the case. The Shaybani conquest of Middle Asia, in the words of a local historian, took place 'without robbery and violence' and led to no essential change in the socioeconomic structure. Only recurrent changes in the composition of the ruling classes took place, and large landed properties were redistributed, so that land was now concentrated in the hands of the ruling house and the elite of the Uzbek tribes. The poet Vasifi wrote that, during the conquest of Herat, one Uzbek emir thought that he and his fellow-traveller were the peasants and explained to them that nothing more than a change of landowners had taken place. The nomadic tribes which accompanied Shaybani Khan occupied the best pastures and crowded the local semi-nomadic and partly even sedentary population in Middle Asia. But some of them later on became sedentary while others transferred to a semi-nomadic economy (Gafurov, 2ft I

Nomads

and the state

1972:528ff.). The lack of grazing territory in Maveraunnahr was an obstacle to any purely nomadic economy there, and the shortage of arable land hindered sedentarization. Although from the very beginning the state created by Shaybani Khan had developed as the second type of state, it contained specific features owing to the fact that some nomads in it were unable and unwilling to adopt a sedentary way of life. The predominance of the appanage system, an ancient tradition in nomadic states, meant that internecine wars continued without respite. As a result the heads of the Uzbek tribes, who in the sixteenth century had formerly been the deputies of the Khan in various sedentary areas and had owned large iqta, in the seventeenth century turned into petty independent rulers (P.P. Ivanov, 1958:68-9; Akhmedov, 1979:22). A situation arose which was somewhat reminiscent of Arabia in the nineteenth century (see pp. 275ff.). These rulers themselves ceased to be nomads and ruled both the sedentary and nomadic population. But since the latter provided their military support, they emphasized their nomadic origin, and maintained links with the nomadic milieu. Thus the social positions of the nomads were much higher than those of the peasants. I think that I have cited sufficient examples to show the basic tendencies in the emergence and transformation of statehood amongst the nomads of the Eurasian steppes. In order to cross the boundary which separated early states from more developed ones, nomadic society had to be integrated (at least in part) with sedentary society into a single social, economic and political system. As a result a nomadic state became a state which, although it had been created by nomads, had an agricultural and urban economic basis. The leading estate of nomads became the ruling class in a sedentary society, or one of the ruling classes of the society. The situation of ordinary nomads in such states was not so consistent. Nomads sometimes paid taxes to the state, but grazing remained their collective property; at least it did so in practice, whereas the peasants did not enjoy this right. It is no coincidence that even the taxes, where they did get paid, as a rule were taken from livestock, and did not involve the utilization of grazing. Within one and the same state created by nomads, the positions of different groups and units of nomads could vary. But the very fact that the nomads in it did not make up a separate society and at best were a more or less autonomous sub-society could not but affect their positions. From the social point of view their positions were better than those of the peasantry, and nomads never joined together with the latter to form a single dependent class. But the sociopolitical differences between the nomadic aristocracy and ordinary nomads tended to become greater, and their interests to become more sharply divided. 262

The Middle

East

In the Eurasian steppes over the course of almost three thousand years two basic types of nomadic statehood may be singled out. Of course, all of the states here had their own individual features and characteristics depending on the circumstances in which they emerged, sociopolitical specificity, relations with the sedentary world and the nature of the latter and, finally, simple historical fortunes. There never was, nor could there have been complete regularity and uniformity. But on the whole the development of nomadic statehood oscillated between the two courses which have been described. Transition from the first course to the second happened far more frequently than it did from the second course to the first. However, development along the second course was rarely completed. In the states created by the nomads of the Eurasian steppes there was always a nomadic hinterland. The nomads in this hinterland frequently partly repeated the cycle of development which already had been completed by their predecessors. This, together with a number of other reasons, explains not only the reversibility of sociopolitical processes amongst nomads, but also the reversible and transitory nature of their statehood. However, the process was still not entirely cyclical. Sedentary states on the borders of the Eurasian steppes, including those created by nomads, in the period between ancient and modern times, developed and became more and more powerful although this broken, multilinear process should not be regarded as a smooth and unilinear evolution. Their economy and technology did not stand still. But in the economic respect nomads changed very little. The only way they could meet a challenge was by extending and strengthening their sociopolitical organization and it is this process, again interrupted and reversible, that we observe when we compare the Scythians and Hsiung-nu with the ancient Turks, and the ancient Turks with the Mongols. It was only in the modern period that the strength of nomads could no longer be compared with that of sedentary peoples. The Middle East I have already noted a number of the characteristics of this region: the absence of any clear ecological demarcation between nomads and agriculturalists, absence of any serious obstacles to trade between the two, a periodic influx of nomads from outside and the frequent weakness of state power. For these reasons, in the states created by nomads the second of the two tendencies, although not always and not fully, tended to predominate. Back in antiquity the Middle Asian nomads, the Parni (Justin, XLI, 4-5; cf. Arrian, History of Parthia, fr. 1 - Fragmenta Historicorum Graecorum, u:586-7), look Iran from the Selcucids and began gradually to settle there 263

Nomads

and the state

(at least their aristocracy did), and merged with the Parthians who were ethnically similar to them. The old Iranian ruling class which had existed since Achaemenid times was forced to make room for them, but it was not exterminated. The Greek and Semitic poleis long retained considerable autonomy. Social and economic relations in Parthian Iran were determined by conditions which had taken shape in sedentary and urban territories with marked local differences. However, old nomadic traditions (for example, the notion that the state as a whole is the property of the ruling royal clan) made themselves felt and links with nomads were not lost. Strabo noted that the Parthians' way of life and customs were in many ways barbaric and Scythian, favouring supremacy and military successes (Strabo, XI.9.2). Parthian Iran was a state with a dynasty of nomadic origin and several ruling classes, the most privileged position amongst them being occupied by the descendants of the Parni nomadic aristocracy. However, the conflicts between the latter and the other ruling classes were never fully eliminated. The Parthian dynasty remained alien in Iran. Neither were the Parthians able to destroy the numerous vassal kingdoms on the territory of Iran. They had to be content with the establishment of their sovereignty over these kingdoms. In the end all of this led to the fall of Parthian supremacy (Ghirshman, 1954:262-6; Neusner, 1963; Istoriia Irana, 1977:100-2). Chronologically the next conquest of Iran by nomads from the Eurasian steppes was that of the Saljuqs. It differed from the Mongol invasions in that it was neither premeditated, nor thought out and planned in advance. It was a chain reaction of events and the unexpected weakness of the Middle Eastern states which led to the creation of the Saljuq Empire. As a result of their conquest of Iran and Asia Minor, together with the adjacent countries, the Saljuqs became creators of an Empire (something which they themselves had hardly expected); the majority of people in this Empire were sedentaries with well established social and political systems and the traditions of statehood which had lasted there for thousands of years (on the Saljuq states see Barthold, 1963; Gordlevsky, 1960; Lambton, 1953; Lambton, 1968; Bosworth, 1968; Cahen, 1968; Agadzhanov, 1969; Agadzhanov, 1973). The conquest and the wars connected with it destroyed many agricultural areas. The nomads took livestock from the peasants and laid waste fields. As usual conquest altered the composition of the ruling classes and led to the redistribution of available land. At a time when the Saljuq invasion was still pending, land-values in Khurasan had begun to depreciate (Baikhaki, 1959:739-41). The Turkic nomadic aristocracy were very obviously pressing the old, sedentary, primarily Iranian aristocracy. But the first sultans already tried to restore the agricultural economy. It was not that the Saljuqs completely changed local ways, rather they 264

The Middle

East

modified them and supplemented them with certain traditions, which went back to Turkic nomadic polities and states, and which were now reinterpreted (Lambton, 1953:60-1; Cahen, 1968:36ff.; Gordlevsky, 1960:77ff.). Personal ties predominated over land ties. Thus, the state was regarded as the corporative property of the ruling clan and the ruling clan was considered to be the executor of the interests of all the nomads (Lambton, 1968:218), although in reality this was not actually the case. Nevertheless, the institution of iqta was far more widespread than it had been under the Buyids (Barthold, 1963:369-70; Lambton, 1965:358f.), that is, there were more allotments of land from which the new ruling class could extract taxes and dues without immediately breaking off their nomadic way of life. It was under the Saljuqs that, for the first time in the Middle East, iqta was used as a means with which the nomadic Turkic aristocracy could establish its supremacy over the Iranian peasantry. This institution was the key one in the introduction of military reform. During the conquests, the dynasty for the most part had relied on tribal corps. Now what was essentially a new army was created, one which had to fight for land rewards. In so far as the distribution of iqta depended on the army commanders, the emirs, the strength of the latter increased and the dependence on them of warrior-nomads grew stronger (Nizam al-Mulk, 1949:129). But the most interesting fact is that not only Oghuz and Turkmen were enlisted in this army, but Arabs and Kurds as well. The dynasty was trying to rid itself of what it regarded as the excessive dependence on its own fellow-tribesmen (Bosworth, 1968:197). However, this did not solve any problems, it only created new complications. Apart from those allotments of land, the revenue from which served to feed the army, vast rewards of land were distributed as iqta or mulk (hereditary rewards) to members of the ruling clan and to the Saljuq elite, and sometimes just as acknowledgement of a position which had already been established de facto. After this the latter began to work for separatism. In the struggle with them the Saljuq sultans began to lean on the Iranian bureaucracy and the Muslim clergy. But those members of the ruling classes in the Saljuq state who favoured its centrifugal tendencies and organized internecine wars knew that they could rely on the support of discontented nomads (Bosworth, 1968:197). The Oghuz and Turkmenian nomads always remained one of the main and most troublesome internal problems facing the Saljuq sultans (Barthold, 1963:372); the problem emerged in the Great Saljuq period, but was felt most acutely in the following period of Saljuq history. Nomads did not forget that they had entered the country as conquerors and they did not want to pay taxes and submit to the governing Iranian bureaucracy in just the same way as the great mass of the population in the Saljuq state did. 265

Nomads

and the state

The Saljuq sultans liked to regard themselves as the shepherds of their people. But to those genuinely involved in the pasturing of sheep the sultans looked like poor shepherds. However, in common with the Turkish sultans of a later period, the Saljuq sultans long adhered to many of the traditions of the nomadic way of life (Gordlevsky, 1960:74; Cahen, 1968:36). But they were leaders of a state in which social and economic relations, and partly even the political organization were determined by people who were not nomads, but agriculturalists and townsmen. Once they began to rely on a new social foundation, the Saljuqs sought to transfer nomads to the level of ordinary subjects and to draw them into the social system which prevailed in the conquered countries. This state of affairs did not pass unnoticed. Al-Idrisi wrote about the Turks from the position of a scholar townsman when he described conflicts between the nomadic dynasty, which ruled a sedentary country, and the ordinary mass of nomads: 'Their princes are bellicose, far-sighted, firm and just, and are distinguished by superb qualities; the people are cruel, wild, coarse and ignorant' (quoted in Barthold, 1963:368). In the Great Saljuq period the ruling powers still somehow managed to cope with their nomadic subjects. Already under Malik-Shah (Bosworth, 1968:79; Lambton, 1968:249) they were forced to pay taxes of a set number of head of livestock, and for the use of pasture. Furthermore, nomads had military obligations. In the eleventh century they continued to play an important role in the army, but efforts were made to use them on the borders of the state, in raids or in the struggle with Christian states. The government consciously split up, divided territorially and dispersed the nomadic tribes, and it encouraged them to become sedentary. Nizam al-Mulk, the most influential Iranian adviser to the sultans Alp-Arslan and Malik Shah, understood the risks which this policy involved. Recognizing that the Oghuz and Turkmen were legally entitled to a privileged position in the state, he suggested that their military detachments be assigned the same positions in the Saljuq state as in earlier states had been occupied by the guard of bought slaves and mercenaries (Nizam al-Mulk, 1949:110). But, although the suggestion of Nizam al-Mulk was only a palliative, his warning that due account should be taken of the Oghuz and Turkmen was forgotten after his death. Nizam al-Mulk was less of a realist than Rashid al-Din, and in the conditions taking shape in the state his internal political programme was doomed to failure. Indeed, it can hardly be said that, during the period in which the single Saljuq Empire was disintegrating and the central power weakening, a single and consistent policy was pursued with regard to nomads, particularly since the influx of nomads from Middle Asia continued. On the one hand the Oghuz-Turkmenian tribes, which willingly participated in internecine wars, were tending to be less and less obedient to 266

The Middle

East

the central powers. On the other hand, governments not only made sure that they paid their taxes regularly, they also made them answerable to deputies in the corresponding provinces or directly to the sultan, while special officials with wide authority were appointed over the traditional chiefs. The dual policy of sultan Sanjar is indicative. He imposed taxes on nomads (Kurpalidis, 1979:106) and tried to place them under the control of the Iranian bureaucracy, while at the same time telling the latter not to push the nomads too hard (Materialy po istorii turkmen i Turkmenistana, 1939:314-15). This policy was very unsuccessful. The response of those Oghuz who had been allotted lands in the Balkh district to excessive extortion and attempts to increase their dependence was, in 1153, to revolt. The Sultan's army was destroyed, the Sultan himself imprisoned and several of the towns of Khurasan plundered. For three years the Oghuz took Sanjar everywhere with them and during the day showed him outward respect, while at night they locked him in an iron cage (Barthold, 1963:393). Under the Saljuqs things did not go as far as they did later on under the Ottomans when the nomads living in the countries conquered by their ancestors became the humble strata of the population. It was in this latter period that the word 'Turk' in the sense of nomad, which at one time had been pronounced either with pride or fear, became an almost disdainful nickname, a synonym of plebeian. Moreover, by an irony of fate this nickname was used by the descendants of those very same Turks, descendants who had only just severed their link with nomadism. Nevertheless, even much later on, in the Turkmenian tradition the notion that the Saljuqs as a dynasty had betrayed their people lived on. 'The Saljuqs announced: "We are Turkmen, we are [your] brothers, but they were of no benefit either to il, or the people"' (Abu-l-Gazi, 1958:70). The Saljuq state (states, to be more exact) without any reservation can be regarded as states of the second type. However, the typical development of such states, in common with that of other states created on the territory of Iran in later periods, was nowhere near completed. Only in the most long-lived of the Saljuq states, the Sultanate of Rum and its successor, Ottoman Turkey, was the development completed. But in the state of the Great Saljuqs and the Sultanate of Sanjar the position of nomads in a sedentary state was not yet established. At one extreme was the dynasty and the ruling class, consisting of nomads and/or former nomads; at the other were ordinary Oghuz and Turkmen who were not part of the ruling class and who pay taxes and were under the obligation to provide certain other dues. In between the two was an aristocracy of individual nomadic groups. As a whole both the social and economic position of ordinary nomads were better than those of the peasants who comprised the basic dependent 267

Nomads

and the state

and exploited class. But the nomads themselves did not comprise either a single class or a single estate. Some nomads turned out to be in a more privileged position than others. Those nomads who were attached to the military estate occupied a higher position than those who made up the tribal corps. Those who served on frontiers received lands (uj) and tax immunity (Guseinov 1979:34). Finally, individual representatives and groups of the ruling classes, attempting to cultivate the support of nomads, gave specific backing to the latter. In contrast to the peasantry, nomads maintained vertical social ties which prevented their situation from becoming any worse. A similar situation was also preserved in many of the states which were created later on by nomads on the territory of Iran. I have already specified the reasons for which the description of the relations existing in the state of the Il-Khans does not belong to the section that it should (see p. 251, n.7). The destruction of the Hulagiis state coincided with the rise of Turkey in the West and of Ming China in the Far East, both of which were sedentary rather than nomadic states. In Iran things were somewhat different. After the fall of Il-Khan power (1258-1353) for one and a half centuries one short-lived nomadic state was replaced by another such state on the territory of Iran (on these states see Petrushevsky, 1949b; Petrushevsky, 1949c; Lambton, 1953). But as has rightly been pointed out by Lambton (1953:175): 'Broadly speaking, the rise of new dynasties did not materially alter the structure of landowning classes though it altered their personnel and to some extent the relative importance of the different sections of the landowning class.' The states of the Chobanids (1340-57), the Jalayrids (1358-1410), the Qara Qouynlu (1410-68) and the Aq-Qouynlu (1468-1501) were created by nomads and headed by different nomadic dynasties. With regard to the basic dependent and exploited class in Iranian society, the peasantry, the policies of these states and individual rulers fluctuated between, on the one hand, the situation which had existed at the outset in the Saljuq state and under the first Hulagiis and, on the other hand, that which had existed under the Great Saljuqs and Il-Khans after the reforms of Ghazan Khan. The first tendency which involved excessive plunder of the peasantry and the monopolization of power in the hands of the nomadic military aristocracy predominated. From the beginning of the fourteenth century iqta were gradually replaced by soyurghals, land rewards which entailed not only tax exemption, but also administrative and legal immunity in exchange for the obligation of military service (Farzaliev, 1979). With regard to the privileged estates in the sedentary population, what had to be decided was what part of them, and on what conditions should this part be admitted into the ruling class, which basically consisted of the 268

The Middle

East

nomadic aristocracy. Local landowners, the bureaucracy and Muslim clergy played only a secondary and subordinate role. Uzun Hasan and his son Yaqub tried to enlist the support of the clergy and bureaucracy as a counterbalance to the nomadic elite (Woods, 1976:7, 285-6). They tried to limit the increase in soyurghal, establish fixed payments of raiyyat in the state of Aq-Qouynlu and prevent the introduction of new payments, but the administrative-financial reform brought only temporary relief. Yaqub died in suspicious circumstances. After his death the emirs dealt with all those who had supported the reforms with great severity (Efendiev, 1979). Ahmad Aq-Qouynlu in 1497 attempted to follow the example of the Turkish sultans to enlist the support of the sedentary strata of the population and cut taxes, he struggled to destroy the soyurghals and opposed the nomadic aristocracy, but he held power for only seven months (Petrushevsky, 1949a; Woods, 1976:291). In Iran a dual position for nomads - in the nomadic polity proper (il, el) protecting their interests, and only by proxy in the state as a whole - was finally established. Ordinary nomads only had to pay low taxes. They were not so much a subject group as a military estate; besides they served in their own tribal corps (khoshun). In comparison to Saljuq times, their vertical social ties were much greater. This was because the nomadic aristocracy now valued its position in the nomadic sub-society more than it had done in the past and did everything it could to strengthen this position while at the same time conserving the relations which had taken shape. In this way, while the encapsulation of nomads in Iran in the Middle Ages increased the power of their leaders, it did not by any means always result in the same consequences as it does today. There was then a greater conflict between the nomad leaders and the central government, and the balance of power was less certain. The position of the nomadic aristocracy on the territory of Iran was not the same as that of the ordinary nomads there. The nomadic aristocracy at the same time acted in two different capacities. Firstly, this was as a distinct group in the ranks of the heterogeneous ruling class in the state, or as one of its ruling classes; and, secondly, as the leading estate of its own nomadic polity, the representative of state power amongst the nomads and the spokesman of the corporative interests of its own polity before the state. In effect, it was through acting in this dual capacity that the nomadic aristocracy best guaranteed its own interests (Petrushevsky, 1949:75-6; Helfgott, 1977:49f.). The position of nomads only changed in any essential way under the Safavids, and then this did not happen immediately. The new dynasty of the descendants of the Ardabil shaykhs was not a nomadic one, but its state was created with the assistance of nomads. The main bastion of support to the Safavids for most of the fifteenth century was a group of nomadic tribes of 269

Nomads

and the state

Azerbaijanian origin, the Qizilbash. Admittedly, already Ismail I tried to limit the oppression of the raiyyats by the nomadic elite (Petrushevsky, 1949c:233), and Shah Tahmasp began to separate and scatter the nomadic tribes (Lambton, 1953:107). Nevertheless, the predominant positions in the state were occupied by the nomadic aristocracy. Only after the reforms of Shah Abbas I the Great did that predominance temporarily come to an end. The creation of a permanent army lessened the significance of tribal corps. Soyurghal ceased to be the main form of land rewards to the nomadic aristocracy and was replaced by tiul, conditional land ownership linked to service; this kind of fief could neither be sold, nor passed on as a legacy. Through this the Safavids tried to return to the practice which had existed under the first Saljuqs. The tax burden on nomads was increased and the most rebellious tribes were physically destroyed. Despite the fact that under the Safavids the system of the indirect government of nomads continued, the dependence of the basic mass of them on the sedentary state and its government increased. Although in the sixteenth and seventeenth centuries the position of nomads was still better than that of the peasants, it was gradually worsening. Thus, under Shah Husayn their taxes were doubled. The dual position of nomads in the state began to grow weaker, and the majority of them came to form one of the dependent and subject classes in the state. At the end of the sixteenth century the nomadic elite found itself in opposition to the Safavids. Although such a tendency existed in later periods, it was not always consistently carried through. In some periods the recurrent development once again became active. The social position of iliyyats (of ordinary nomads, that is) was, as before, higher than that of the raiyyats. Nadir Shah, like the previous rulers of Iran, used the policy of indirect rule of nomads. Nevertheless, despite his Qizilbash origin, he tried to weaken the nomadic aristocracy to a greater extent than his predecessors had done, and he broke up and scattered those tribes which were hostile to him (Lambton, 1953:131). At the same time individual Khurasan tribes, primarily the Afshars, were his main support (Arunova and Ashrafian, 1958:111-12). Fath Ali Shah, founder of the Qajar dynasty, looked upon his own tribe as the mainstay of his dynasty and Iran as the conquered country. Under the first Qajars the position of nomads in the state was in many ways similar to that of nomads in the preceding Safavid period (Lambton, 1953:135-62). The ruling class of landowners in the country included the nomadic tribal leaders. The latter divided out amongst their fellow tribesmen the lands which had been especially set aside for them and were responsible for the collection of taxes in these lands. As was the case under previous dynasties, some nomadic tribes became 270

The Middle

East

weaker, whilst others grew stronger and privileged tribes paid no taxes at all. As the nomads in the army travelled round the country they devastated the lands of agriculturalists as they went, and the government took no effective measures to put a stop to this. Even in the second half of the nineteenth century when the policy of the Qajars towards nomads changed, control of the tribes and the collection of taxes from them remained a serious problem. Only in the present century and with the assistance of extreme and severe measures were the Pahlavi basically able to put an end to the distinct status of nomads and relegate the vast majority of them to the position of an unprivileged class which, because groups of nomads were transferring to agriculture and other occupations, was being eroded both socially, and economically. Whether, from the political point of view, the problem of nomads in Iran has finally been settled only the not-too-distant future will tell. The problem of 'nomads and the state' was somewhat different in neighbouring Afghanistan (see Reisner, 1954; Masson and Romodin, 1965; Gankovsky, 1958; Gankovsky, 1958a; Lockhart, 1958; Gubar, 1959). However, amongst the different types of nomadic statehood in the Middle East, the 'Afghan model' is the one which generally is not very distinctive. The main difference consists in the fact that during the resettlement of the Afghan tribes in the fourteenth, fifteenth, sixteenth, seventeenth and eighteenth centuries, some of the tribes conquered cultivated territories and their populations and thus became landowners before the creation of a state proper and without such a state. Migrations and appropriation were facilitated by the fact that the agricultural Iranian-Tadjik and Indo-Aryan population in the broad expanse of territory between India and the Hindu Kush was partially wiped out during the Mongol invasion and Timur's campaigns and those who did survive retreated to some extent to inhabit less dangerous places. Thus, once again the view that not every conquest of agriculturalists by nomads leads automatically to the emergence of a state is substantiated. Wherever the agricultural population is relatively small and dispersed, it has no strong political organization, and its subjugation and the exploitation of it do not demand any essentially new forms of organization or management from nomads, a nomadic state does not necessarily emerge, at least not immediately. The Afghan tribal polities were no more than chiefdoms. The subjugation of the agricultural population was only one of the factors which in the long term led to an increase of social differentiation and the emergence of the state amongst the Afghans. Another factor was their own gradual sedentarization and the ensuing consequences. A third was their active participation over many years in the political life of the Middle East; their dependence in the sixteenth and seventeenth centuries on the Safavids 271

Nomads

and the state

and Moghuls enabled the Afghan khans to strengthen their influence amongst their fellow tribesmen (see p. 215), but at the same time it showed them the benefits to be derived from the creation of their own state. The Afghans played an active role in Babur's conquest of India, and then in the prolonged struggle between the Safavids and Moghuls for Qandahar. The opportune moment arose for the Afghan aristocracy at the beginning of the eighteenth century when India, Iran and Middle Asia were in a state of decline and political disintegration. And, as was frequently the case in such situations, external expansion forestalled internal consolidation. At the beginning of the eighteenth century the Ghilzais in Qandahar, under the leadership of Mir Wais of the khan khel of Shah Alam, and the Abdali in Herat, under the leadership of representatives of the khan khel of the Saddozais, broke away from their dependence on Iran. At the beginning there was conflict between them. But already by the end of 1721 the son and heir of Mir Wais, Mir Mahmud, leader of the Ghilzais who were still leading a nomadic and semi-nomadic way of life, decided to march on Isfahan. In the following year he conquered a considerable portion of Iran and established himself as shah there. The short-lived state of Mir Mahmud had all the basic characteristics of a nomadic state of the first type. The government had only one bastion of support, the Afghan army which was ruthlessly robbing and devastating the country. Admittedly, the conquerors at first did try to preserve the Safavid administration, but they soon began to repress it once they had discovered its unreliableness. Neither did they manage to get the Iranian nomads onto their side. Nevertheless, the Afghans basically preserved the system of government which had previously existed, although they did increase taxation, quite apart from simple robbery and extracting contributions (Miklukho-Maklai, 1954). The only original feature in all of the measures undertaken by these conquerors is that they could not, not that they even wanted to, include other Afghan tribes in the conquest of Iran. They even contrived very quickly to quarrel amongst themselves. As a result they ceased to receive any support from Afghanistan. Mir Wais and his successors tried to tax certain Ghilzais khels, but this was met with direct opposition. The old animosity between the Ghilzais and Abdali also had an adverse effect on Mir Mahmud's undertaking. That is why the struggle of the Afghan conquerors against Nadir Shah was doomed to failure. Nadir Shah who conquered almost all Afghanistan pursued a policy of weakening the Ghilzais, particularly the Western Ghilzai, while at the same time he tried to get the Abdali khans onto his side. However, once they had become stronger it was they who were able to create the first national Afghan state. 272

The Middle

East

The state of Ahmad Shah Durrani was also a conquest state. However in it the Abdali (Durrani) tribes, in which nomads were in the minority, occupied the privileged positions. At the same time vassal domains and, even more so, the territories of the other Afghan tribes retained considerable internal autonomy and were ruled according to their own traditions. The percentage of nomads in the tribes in West Afghanistan was particularly high and all that the shah required from these tribes was warriors (Elphinstone, 1819, n: 104, 105). The majority of Afghan tribes did not have to pay taxes, or if they did have to pay them the taxes were low, sometimes they were only purely symbolic (Gankovsky, 1958:69). Ahmad Shah 'has already the penetration to discover that it would require a less exertion to conquer all the neighbouring kingdoms, than to subdue his own countrymen' (Elphinstone, 1819, i:281). The successful campaigns of Ahmad Shah enabled him to lay the basic burden of taxation on the non-Afghan population in the areas which he had conquered. In doing this he used the administrative apparatus inherited from the Safavids, Nadir Shah and the Moghuls. A sober and far-sighted politician, Ahmad Shah limited his conquests in India to the Panjab and did not strive to create a vast empire and move his capital there. He did not want the dynasty to become isolated from the Afghan core of the state and in one of his poems he wrote: T forget about the throne at Delhi when I recall the mountain heights of the country of the Afghans.' At the same time Durrani and other Afghan tribes in the Durrani state continued to seize cultivable lands and centrifugal tendencies quickly developed. The Afghan khans were invested with great authority in the territories they had conquered and they intensified the exploitation of the dependent agriculturalists - hamsaya. However, they were acting not only in their own interests, but also in the interests of their tribe. With regard to their fellow tribesmen the majority of Afghan khans did not wield any compulsive power and they were unable to collect taxes and other dues from these tribesmen. 'I do not know two instances of a Khaun taxing his Oolooss for his benefit; but the regular tax on "Hammsanyas", the infidel tax on Hindoos and customs collected on merchandise passing through the land of the Oolooss are, in several cases, appropriated by the Khaun' (Elphinstone, 1819, i:263-4). But, most importantly, the Durrani state from its very beginning cannot be described as a state which was created by nomads. The latter merely participated in its formation and, particularly, in its conquests from which they derived certain benefits, sometimes in the form of further opportunities and incentives to renounce nomadism (cf. the polities of Baluchistan - see Swidler, 1973; Salzman, 1978a; Pastner, 1978).

Nomads

and the state

The Near East With regard to the Near East as a whole an observation of Gellner (1973:1), although actually referring to a larger region, is apposite. 'The history of the Islamic Middle East can, from its very beginnings, be written to a large extent in terms of the interaction between the nomads and the sedentary and urban population' (cf. Asad, 1973:71). There were other diverse factors which influenced the nomadic statehood of the Near East: the fact that agriculture was concentrated primarily in oases; the poor development of large irrigation systems (excepting in Mesopotamia, Egypt and the ancient Yemen) which would favour centralized government; the division of labour between nomads and semi-nomads; sometimes the aspirations of nomads to monopolize the ownership of stock or of specific species (the Tuareg, nomads in Mauritania, partially the Fulani in the Sudano-Sahelian marginal zone; see Gallois, 1972:301-2); the relatively low density and scattered distribution of the nomadic population in the vast territories which is characteristic of many sub-types of Near Eastern nomadism; and, finally, the important role played by Islam, 'a privileged instrument of expansion and influence' (de Planhol, 1976:126). Nevertheless, the ecological and historical conditions in the various areas of the Near East were so different that nomadic statehood in all these areas had clearly defined local characteristics. The lack of large irrigation systems in Arabia since the Middle Ages and the fact that agricultural and urban life there were linked to oases were always obstacles in the way of the unification of the region into one state. Apart from the ecological obstacles, the economic potential of the country was too low to be any incentive for the emergence of large nomadic polities and for the creation of large and powerful sedentary states. In contrast to the nomads of the Eurasian steppes, in Arabia relatively stable polities of several nomadic units were rare and short-lived. Thus the polity of Nuri al-Shaalan in the first quarter of the twentieth century included the Rwala, Aulad 'Ali and al-Mikhlaf, and certain small non-Anazah tribes, the most powerful amongst these being the Rwala. All these tribes were regarded as equal and independent. Without the consent of their sheikhs Nuri had no right even to declare war. All of the actions which he took to limit the independence of the individual tribes were regarded as the usurping of their legal rights and their success was only relative and short-lived (Musil, 1928:506; Oppenheim, 1939:108). The amount of land in Arabia which was suitable for cultivation was insufficient to allow the mass sedentarization of nomads there in conditions which were beneficial for them. The subjugation by the Bedouin of individual oases usually occurred on a local level (see p. 223). 274

The Near East It is indicative that the first states in pre-Islamic Arabia primarily emerged on the peripheries of the peninsula where nomads were the immediate neighbours of richer and more developed countries. Amongst these states were those of the Kidarites, the Nabataeans, the Kindites, the Himyarites and the Lakhmids. In Arabia itself usually it was sedentaries who encouraged centripetal tendencies and they employed different methods, ranging from the ideological to the purely military, to overcome the separatist tendencies of nomads. It is curious that the situation which emerged in the nomadic regions of Arabia in the sixth century A . D . was somewhat similar to that which existed in Mongolia in the twelfth century. The emergence of the Arab kingdom Palmyra and then of the buffer states of Hira and Ghassan prevented the free movement of nomads to the north; the occupation of South Arabia by the Abyssinians and then, in about 570, by the Persians prevented their migrations to the south. Trading along the Spice Route declined and was much less busy than it had been in antiquity. Between 591 and 640 there were a series of droughts in Arabia (de Planhol, 1970:444). All this led to overpopulation which at first meant that raids and conflicts amongst the nomads themselves were intensified (Bishai, 1968:61-2). Unification under the leadership of Mohammed, who himself had been born into a trading-urban milieu, then followed. Arabia did not possess a Jenghiz Khan. However, Mohammed's state would scarcely have been able to survive for any length of time if it had not embarked on extensive conquests outside Arabia. The first Wahhabi state tried in some ways to follow the path trodden by Mohammed and his immediate successors (Vasilev, 1967). But as things turned out in the intervening centuries the world which surrounded Arabia had changed so much that the Wahhabi state had little chance of prolonged success. Nevertheless, from time to time the subjugation by nomads of a number of sedentary territories resulted in the emergence of small states or state-like polities. It is noteworthy that in some of these nomadic states the first tendency prevailed, while in others it was the second tendency, although even here local characteristics manifested themselves fairly clearly. However, we only have detailed information about these states in the eighteenth and nineteenth centuries. In so far as conditions for mass sedentarization were not suitable and even extra-economic opportunities for exploiting conquered or subjugated territories were limited, a dynasty had to choose between two alternatives. Either it could remain nomadic and, as before, be the leading aristocratic estate in a nomadic society, or it could consolidate its power in the oases and thereby risk losing the support of its Bedouin fellow tribesmen, for the latter would derive little advantage from that latter course. It is not surprising,

Nomads

and the state

therefore, that certain rulers tried to manoeuvre and not to embark on either of the available paths. For example, members of the al-Humaid house of the Banu Khalid owned in the seventeenth and eighteenth centuries, up to the time of the formation of the Wahhabi state, all of al-Hasa and part of Najd and continued to lead a nomadic life, never living in their capital al-Hufuf (Niebuhr, 1780, u:180; Oppenheim, 1952:133ff.). The Shaalans, the Rwala sheikhs, also lived like this when they managed to establish their power for a short period in al-Jauf. However, such a way of creating a state was too unstable and impermanent. In other cases Bedouin sheikhs managed to gain control of oases and towns and then they became the rulers of them (Blunt, 1881, i:260—1, 270), gradually distancing themselves from their own tribes and making themselves the leaders of the sedentary aristocracy. After a certain period they were ready to take up arms to defend their acquired property from the claims of nomads. At the same time they tried to maintain their links with nomads and to preserve kin relations with the Bedouin elite. Such, for example, was the history of the Jabal Shammar emirate (Pershits, 1961:168ff.; Rosenfeld, 1965:81ff.). Abd Allah, founder of the dynasty of the Ibn Rashidids, forced certain non-Shammar nomadic tribes to acknowledge his power, and at first just Shammar tribes were the mainstay of his support. But they were only really interested in pillage and the collection of tribute, consequently their influence began to be limited systematically. Abd Allah's son, Talal, already was subjugating the Bedouin together with those who his father had subdued a generation earlier with the help of the Bedouin (Palgrave, 1866:53). The Rashidid army now basically consisted of slaves, freedmen and mercenaries, and the Bedouin militias were only used in a sup­ plementary capacity. The Rashidid policy towards nomads was fairly ambiguous. It was limited to a general control of inter-tribal relations and the collection of zakat, which was not always successful. The tribal sheikhs and those nomads who were drawn into the military campaigns of the emir were exempt from payment of zakat (Musil, 1918:24-5). The tribes were governed by their own sheikhs, pastures were the collective property of the tribes and the Rashidids almost never interfered in their internal affairs (Doughty, 1888, i:537). The Rashidids did all they could to emphasize their Bedouin connections, primarily to the Shammars and they employed a variety of methods in their attempts to preserve the loyalty of the latter. To the sedentaries of Jabal Shammar the Rashidids were emirs, to the Shammar they were merely sheikhs. They dressed in Bedouin clothes, spent from three to eight or nine months a year in the desert (Huber, 1891:165) and systematically provided the poor Shammars with food (Doughty, 1888, i:6!0). 276

The Near East In Arabia noble Bedouin origins were so very prestigious that even such sedentary rulers as the Saudis, Kuwait al-Sabah and the Bahrain al-Khalifah liked to stress their Bedouin descent (Dickson, 1951:111). On the whole it may be concluded that until the formation of the contemporary state in Arabia nomadic society there remained socially differentiated internally, but not yet a class society. Only where Bedouin society came into contact with sedentary society, where some nomads became sedentary and the nomadic aristocracy became the ruling class over the sedentary population did class differentiation begin to develop. But such development was already the development of a sedentary, not a nomadic society. Even the Wahhabists in the eighteenth and nineteenth centuries, after they had made a number of nomadic tribes dependent on them, could not change fundamentally the social and political organization of those tribes. The evolution of the nomadic societies of the Sahara was different. The uniqueness of Tuareg society (more precisely, of Tuareg societies) consists, amongst other things, in the fact that the submission of one group of nomads to another was, more than in other nomadic societies, the basis of very clear-cut social stratification. The majority of scholars believe that the opposition between the noble Tuareg (Ihaggaren, Imajeren) and their vassals (Imrad, Kel Ulli) is connected to the subjugation by camel-herders of the more ancient goat-herding nomads of the Sahara (Lhote, 1955:200; Birket-Smith, 1957:134; Bourgeot, 1975:19-23; Keenan, 1977:16-18; Gast, 1979:202-3). Thus, according to Nicolaisen (1963:405, 409-10, 436, 479ff.), Tuareg societies emerged on the basis of two different variants of pastoral nomadism. In the first centuries A . D . Berber camel-herders, the Lemta, moved southwards and invaded the territory of the Berber-speaking goat-herders of the Sahara. New camel-herding groups appeared in Tassili, Ahaggar and Ayr as a result of the migrations caused by the Hilalian invasion of North Africa in the eleventh century. These groups are usually identified with the Hawwara Berbers who moved southwards from Fezzan. It is thought that it was these migrations and the subjugation of the more ancient nomads which decisively shaped the division between the noble Tuareg and their vassals. However, Pershits (1969:342-3) has pointed out that haematological research shows that the blood-groups of noble and vassal tribes are identical (Mendoul and Jacquemin, 1953, non vidi). Consequently, in his view there are no grounds for thinking that they belong to different Tuareg ethnic groups. He believes that at the basis of the social differences between the noble Tuareg and their vassals is the division of labour, dictated by ecological circumstances, between camel-herders and goat-herders. The camel-herders were militarily stronger and so in an advantageous position; 277

Nomads

and the state

they were able to monopolize the ownership of camels and even to use major types of weapons. But, in essence, this hypothesis still allows for subjugation. The history of the formation of Tuareg societies was very complex and took place over a long period. For this very reason, therefore, there is no need for one hypothesis to exclude all others. History and anthropology very often can make little use of Occam's razor. The history of the southern Tuareg had long been closely connected with the history of Sudanese societies and states. Admittedly, there are no positive indications that the Berbers (Tuareg) penetrated Hausa areas earlier than the fifteenth century (Smith, 1971:183). But in fifteenth-century Timbuktu and Walata tributes were paid to the Tuareg (Levtzion, 1971:140). Thus, I shall deal with the more 'distinctive' example of the northern Tuareg, the Kel Ahaggar, about whom a great deal has already been written. Can it really be said, as many scholars have claimed is the case, that Kel Ahaggar society was a class society and had a state? The great social differences between the two basic estates in that society, the Ihaggaren and the Imrad, are very evident. 'The suzerains call the Kel Ulli tameksit, that is to say "sustenance". Nothing could be more unambiguous' (Gast, 1979:206). The Ihaggaren consisted primarily of the military aristocracy and considered itself to be the owner of all the land and to have the right to free movement in the territory of its ettebel (ettebelen in the plural, meaning drum group), which are called chiefdoms by Nicolaisen (1963:217). Their own pastoral (camel-herding) economy was very limited and mainly worked by slaves. Their main sources of revenue were the tributes and obligations of the Imrad, raids, payment for protection and for the passage of caravans through their territory, etc. The Imrad owned most of the livestock and had their own slaves. They paid fixed dues (tiwse) annually both to the amrar, representing the Ihaggaren as a whole in their ettebel, and to individual families or matrilineages for patronage (temazlayt). They pastured the livestock belonging to the Ihaggaren, but they also had usufructuary rights to the livestock and were able to take camels for certain periods for trading purposes. In their turn they had to lend their livestock to the Ihaggaren for temporary use, in extreme circumstances they had to give up without indemnity a number of livestock, to participate in the military undertakings of the Ihaggaren and to share out their booty when they made independent raids (Duveyrier, 1864:273, 334f., 374, 385; Benhazera, 1908:54, 139, 148; Lhote, 1955:194, 371; Nicolaisen, 1963:399-400, 403-5, 436; Keenan, 1977:32f.; Bourgeot, 1979:147ff.; Gast, 1979:205f.). Of the other vassal estates the most important and numerous were the 278

The Near

East

Isekkemaren, who were particularly active in trade (Nicolaisen, 1963:10; Keenan, 1977:28-9). The other dependants in the population, slaves (Iklan) included, had less productive and social significance. Judging from the available data, the Tuareg polity (tegehe) in Ahaggar may be regarded not simply as a society of different estates, but as a class (early class) society. At the same time the primarily collective forms of dependence which were established between the Ihaggaren and Imrad and the elements of reciprocity within them (in the political sphere included, relations about military defence and protection is what I have in mind here) are evidence of incipient forms of undeveloped class relations (Nicolaisen, 1963:404; cf. Bourgeot, 1975:27). The Ihaggaren were the corporative holders of political power. Admittedly, there were also individual leaders, amrars, in the various ettebelen who were chosen from specific matrilineages. There was also a su¬ preme chief, amenukal , who at the same time was amrar of one of the ettebelen. For the Ihaggaren the amrars were usually the military leaders, the organizers of public life and also of the redistribution of surplus considered as their collective usufruct. At the same time in their capacity as chiefs of an entire ettebel the amrars regulated relations between the different estates, organized the distribution of the territory of their ettebel and received special payment for this service. They were also paid by caravans for free passage through the territory of their ettebel and received half of the booty which their vassals procured in their various raids. The amrars were endowed with legal functions, set punishments and imposed fines, which were for their own use, and saw that the Imrar fulfilled their obligations. Part of these obligations were directly to the amrars (Duveyrier, 1864:353, 427-8; Lhote, 1953:193f.; Nicolaisen, 1963:336, 400, 435-6; Keenan, 1977:38-9, 40; Gast, 1979:205-6). The nature of the power of the amenukal was also contradictory, but in a rather different respect. The power of the amrars was a strong counter­ balance to that of the amenukal despite the fact that the amenukal was amrar of a larger ettebel (the Kel Rela amongst the Kel Ahaggar, the Uraren amongst the Kel Agger). The amenukal had definite religious prestige and acted as the representative of his polity as a whole to the outside world; his power increased when his polity acted as a single unit, but in the nineteenth century this happened only rarely. The amenukals had independent sources of revenue which were linked to their office, the major one of which was payment for protection from neighbouring tribes, and they even had three or four assistants (khalifa) (Benhazera, 1908:53, 55; Lhote, 1955:189ff.; Ilriggs, 1960:136; Nicolaisen, 1963:397-8, 400-1, 435; Gast, 1979:206). At the same time, amongst the Tuareg of Ahaggar in fact there was no distinct apparatus of power and the functionaries of such, both of which are characteristic of all states.

Nomads

and the state

The boundary-line between early states, particularly in the first stage of their development, and pre-state units (developed or complex chiefdoms) is always rather tentative and hard to discern. It is extremely difficult to draw this line for such a specific society as that of the Tuareg of Ahaggar which is very small, which is rigidly partitioned into different estates and which has a high estate that at one time monopolizes military affairs and coercion and acts as the corporative owner of key resources and the collective exploiter of the dependent strata. More individualized forms of dependence, temezlayt, in time acquired great significance. However, they were only derivations of the interrelations between different strata of Tuareg society as a whole. Collective tribute was of central significance in this society, clientage secondary. Thus, the society of the Tuareg of Ahaggar can be regarded as a class (early class) society, but still as a pre-state society. Be that as it may it is quite clear that Tuareg society in Ahaggar achieved, in the absence of any significant groups of dependent and exploitable agriculturalists, a level of social inequality which was very high amongst nomads. In this respect the northern Tuareg were very different from the southern Tuareg. Amongst the Kel Geres Tuareg the ratio of noble to dependent agriculturalists was 1:10 (Bonte, 1979:183). But in Ahaggar the tenants-metayers (Izeggaren, Harratin in Arabic) only began to settle in the second half of the nineteenth century, and towards the end of the century the settled population numbered only a few hundred (Keenan, 1977:100). Such a situation is unique amongst nomads and requires some explanation. Frequently the explanation is to be found in the particularities of the relations between the Ihaggaren and Imrad. It would appear that nowhere else did the dependence of certain groups of nomads on others within one single economic and sociopolitical system become as great as it did amongst the northern Tuareg. But this was possible not only for the ecological reasons which prompted economic specialization and the military-political specialization of the.Ihaggaren. The situation which took shape was also in many ways beneficial for the Imrad. They had military protection and concentrated all the basic forms of economic activity, including trade from the end of the nineteenth century, in their own hands. The obligations of the vassals were not too much of a burden to them. There is no data to indicate that there were any really serious conflicts between the Imrad and Ihaggaren. Furthermore, it is no coincidence that many vassals were richer than their noble suzerains (Duveyrier, 1864:220, 354, 355; Benhazera, 1908:151f., 160-1). However, I should like to point out another factor which is linked to the sociopolitical changes which took place amongst the Tuareg of Ahaggar, for their society was certainly not static and slow to change. Up to 1660 all the northern Tuareg were part of a vast state unit headed by sultans from the 280

The Near

East 9

Imenan tribe which regarded itself as shurfa. We can only guess what influence this exercised on the sociopolitical organization of the Tuareg. However, characteristic of the last period before French rule is a gradual weakening of their political organization and an improvement in the position of the Imrad in Tuareg society. After the uprising under the Uraren, the northern Tuareg separated into the independent polities of Kel Agger and Kel Ahaggar. At first all or the majority of vassals were directly subject to the amenukal and to only one of the noble groups, the Tegehe-n-ou-Sidi. But already in the middle of the eighteenth century the amenukal Sidi ag Muhammed el Khir had to give way to the demands of other noble groups. The political organization of the Tuareg of Ahaggar and particularly the degree of centralization amongst them once again became weaker. At first three ettebelen, headed by the Ihaggaren Kel Rela, Taitok and Tegehe Mellet, emerged. Each of them had their own vassals and enjoyed a considerable extent of autonomy. The strongest ettebel and the one which had the most vassals was the Kel Rela, and on this account there was constant discord between the three of them. Furthermore, during the eighteenth and nineteenth centuries the Tuareg of Ahaggar were frequently at war with the Tuareg of Ayr and their other neighbours (Nicolaisen, 1963:393f.; Keenan, 1977:25-32, 53f.; Gast, 1978). In the nineteenth century the power of the amenukal continued to diminish and they even had difficulty in maintaining order within their own ettebel (Nicolaisen, 1963:397-8, 401). As the internal discord and external wars continued the social position of the Imrad improved, slowly but surely. According to Keenan (1977:55-6) this process began as early as the second half of the eighteenth century. Amongst the Ihaggaren the need for supplementary military forces emerged. Right up to the time of the arrival of the French in Ahaggar, vassals had the right to own camels and the same weapons as the noble Tuareg. Towards the end of the nineteenth century (Keenan, 1977:57-61), or even earlier (Duveyrier, 1864:331; Nicolaisen, 1963:398, 435), vassals had a considerable say in the election of amrars and amenukal. The Taitok Ihaggaren even began to marry their Imrad (Briggs, 1960:136; Nicolaisen, 1963:402-3). Several Ihaggaren families did not so much order as ask their vassal to make traditional payments (Lhote, 1955:225). It may be surmised that at the time when the French arrived in Ahaggar Tuareg society there was on the brink of decisive change. When Keenan (1977:11) characterises these changes as 'class revolution' he is evidently exaggerating. However, they could indeed have brought about the fall or the disintegration of the power of the Ihaggaren and so eased or obliterated " Shurfa are descendants of the prophet Muhammad. The term literally means 'noble'. Normally shurfa lineages claim special political and religious status.

281

Nomads

and the state

the dependent status of the Imrad. It is doubtful that Tuareg society in Ahaggar would then have remained even an early class society. In Tuareg society in Ahaggar the third tendency which is linked to the emergence and transformation of nomadic states, division of labour that coincides with social stratification within the society, is to be observed. This tendency remains somewhat unclear amongst the Tuareg, partly because the state itself had not formed, partly because of other concomitant tendencies and, mainly, because the division of labour was between the pastoralists themselves and not between pastoralists and agriculturalists. In certain other Tuareg societies, however, this third tendency is manifested more clearly. Pershits (1968:351) correctly stressed the lack of political development amongst the Tuareg: what made it impossible for the amenukals to consolidate their power over the nomadic tribes included the fact that the population was thin on the ground, and, more significant still, that neither the Kel Ahaggar not the Kel Aj jer possessed even a single large oasis which could serve as a base for such strengthening of their power. This idea can be taken further. The fact that so few agriculturalists were available, who might have been exploited, once again reversed the tendency towards social differentiation amongst the Tuareg of Ahaggar, and so, notwithstanding the oddities of the region and of its sociopolitical organization, the Tuareg returned to a pre-class stage of development. So in the end, the general laws of the development of nomadic societies prevailed over local idiosyncrasies, even in this unusual case. There were many distinctive aspects to the problem of nomadic statehood in North Africa (the Maghrib). The nomads and agriculturalists in this region lived lives which were far from distinct, and the borders between the nomadic and sedentary world in the Middle Ages were very flexible (de Planhol, 1968:140ff.). Even if the nomads did migrate to areas near the Sahara for the winter, their summer camps were directly adjacent to arable lands, for example in Tell and the surrounding areas. Furthermore, it was only the camel-herding groups which migrated to the South; sheep-herders did not participate in winter migrations (Dhina, 1956:421—4). Besides, nomads were always drawn to the wide open plains of the Maghrib, and it was not only from Asia that nomads migrated to North Africa. Unfavourable conditions, droughts and other factors drove the nomads of the Sahara to the fertile lands adjacent to the sea. Sometimes this involved conquest, sometimes gradual infiltration, and sometimes they were invited by sultans wishing to set one group of nomads against another. Frequently nomads appropriated land in the process and placed agricultur­ alists in a dependent position (Dunn, 1973:85, 92-3). The nomads of North Africa had every reason to preserve their situation as it was; this was because over the course of many centuries they had been 282

The Near East the ruling military force there and consequently had enjoyed the corresponding opportunities to exploit agriculturalists and townsmen. During the Middle Ages only very few of them became sedentarized (de Planhol, 1968:152). All of these factors could not completely preclude, but they did essentially limit the manifestation of the second tendency in the evolution of nomadic states. Another distinctive feature of North Africa was the situation of sedentary societies there. Not infrequently in the Eurasian steppes, and sometimes even in the Middle East, nomads were opposed by large and powerful states. As a rule, the subjugation, quite apart from direct conquest, of these states by nomads was in any form possible only when the states were weak and decentralized and/or nomads themselves previously had been unified and had had a centralized government. In North Africa the Hilalian invasion of the eleventh century, which Ibn Khaldun called 'the plague of locusts', struck a powerful blow at local sedentary statehood. Admittedly, the point is now sometimes made that the destruction and ruin caused by the invasion has been exaggerated by Ibn Khaldun and the French historians of the old school who followed him (see, for example, Fischer, 1977:244-5). But the fact remains that from the eleventh century onwards nomads were a very severe and constant problem for all the states of North Africa. Sometimes the North African states were so weak that nomads were able to rob and exploit the sedentary population, while they themselves never centralized beyond the tribal level. Even the largest and strongest states could not do without their support, were unable to drive the nomads out of their territory and, therefore, were forced to make various concessions for them. For these reasons there was not always any essential need in North Africa for nomadic statehood of the first type to emerge or, even more so, to exist for any extended period. Thus the actual opposition between nomadic and sedentary statehood in North Africa was considerably less marked than it was in the Middle East and even less marked than in the Eurasian steppes. The social and political criteria were more fluid and less defined. It could be that with regard to certain periods in the history of North Africa it is more apposite not to talk about opposition between sedentary and nomadic statehood, but about a statehood in general and its characteristics, particularly the role of nomads in the emergence, functioning and replacement of certain states by others. While they did not always feel a need to create their own states, nomads 10

111

In the history of North Africa this question is part of the wider problem about the interrelations between states, on the one hand, and tribes with segmentary organizations, on the other. Amongst the latter there were also agriculturalists whose social organization, similar to that of nomads, probably emerged under the influence of the latter by a kind of chain reaction. The role of these agriculturalists in the creation and replacement of certain states by others little differed from that of nomads.

283

Nomads

and the state

nevertheless played an extremely active role in everything connected with states (on this point see Ibn Haldun, 1925; Marcais, 1913; Brunschvig, 1940; Brunschvig, 1947; Julien, 1956a; Terrasse, 1950; Le Tourneau, 1969; Laroui, 1970; Abun-Nasr, 1971; Levtzion, 1971; Levtzion, 1977; Fischer, 1977). Many states in North Africa were created as the result of conquests. In effect, it was nomads who were the conquerors and who frequently provided the new states with dynasties. Fairly quickly, however, members of the dynasties moved to the towns together with their closest associates and stood at the head of sedentary populations. But most of the nomads remained such and felt neither the desire, nor the need to become sedentary. A contradiction emerged. The dynasty and its government strove for centralization and the regulation of state life, and in order to accomplish these effectively they had to subjugate nomads or at any rate to control them. Nomads were only interested in the best conditions for exploiting sedentary populations and in preserving their own practical independence of any government which they were prepared to tolerate and even to support in so far as the policies of the latter corresponded to their interests. The situation was a little different in those instances where by origin a dynasty belonged to a sedentary religious milieu. The weakness of state power, specific ecological circumstances, ethnic differences and a number of other circumstances impeded the creation of large nomadic polities which were necessary for large-scale conquests. On the other hand the ascendency of Islam and the marabout cult, which was very widespread in North Africa, created a climate favourable to the activities of all kinds of saintly mediators and reformers; such individuals were able to unite nomads for a period to fight under the banner of religious purity. But even in such states as those, nomads remained an autonomous and scarcely controlled force, constantly seeking more privileges and new opportunities to exploit the sedentary population. The almost continuous internal dynastic disputes in the different states of North Africa also helped them to advance their aims. 'The new sultan will do well to beware of breaking the bonds which unite him with the comrades of his former life. It is they who will provide him with the military strength of which he will be in need to keep his throne. It is they who will supply the army with that so indispensible nomad contingent which, in modern Morocco, is called the "guish" (jaysh): they will be the makhzan tribesmen of the naissant empire' (Marcais, 1913:245). In these conditions different governments had only two possible political principles for nomads at their disposal, external expansion and the old 'divide et impera'. One of these did not necessarily preclude the other. Gellner (1969:3-4) has pointed out that in medieval Moroccan states 284

The Near

East

there were three groups of tribes: the privileged 'inner circle' of tribes, the main support of the dynasty (guish) in whose favour taxes were paid; the 'middle circle' who paid taxes; and the 'outer circle', the tribes refusing to pay taxes. 'The history of Morocco can be written as the story of the struggle of successive dynasties to maintain their power and authority. The two main problems facing them were the recruitment of the Guish and the holding or expanding of the land of siba.' This is the model which, with certain modifications, was accepted by the French in the colonial period when several tribes, basically Berber ones, were accorded more privileges than were others. The privileged tribes of the states of North Africa paid no taxes and regarded the sedentary population merely as an object of exploitation. As representatives of these tribes their sheikhs received large financial subsidies from the state, as well as fiscal and land rewards for military service. Even when nomads migrated to the south they left tax collectors behind to collect the dues which their raiyyats owed to them. In this way ordinary nomads also received their share of the revenue derived from the exploitation of the sedentary population (Marcais, 1913:253ff.; Brunschvig, 1947:100ff.). Social differentiation within the nomadic tribes, particularly in the guish tribes, was still at the pre-class stage. However, with regard to the sedentary population these tribes were a privileged estate in a class society, and the tribal aristocracy was part of this society's ruling classes. The sheikhs of the privileged Makhzen nomadic tribes were connected with the rulers by marriage and they participated in the making of the most important decisions and in the main events in the life of the state (Marcais, 1913:242ff., 258-9). If a government either could not or would not satisfy all the demands made by nomads and if the split between the two went too far, then that government would lose its support in the nomadic milieu. Ibn Khaldun's conclusion that Bedouin states exist only for three generations (Ibn Khaldun, 1967:138ff.) was based on the rich historical example of Morocco. Thus the governments of the different North African states usually tried not to obliterate entirely their dependence on nomads, but to limit it as far as possible, to moderate the tyranny of nomads and to control their actions, making use as they did so of the animosity between the different nomadic tribes. The Almoravid empire emerged as a fairly typical nomadic state of the first type. Behind the religious reform-movement which drew the Sanhaja Berbers from the Sahara to Morocco and Spain, there were specific historical circumstances. 11

" It should be pointed out that not all the guish tribes were nomadic; some were sedentary or cx-nomadie.

285

Nomads

and the state

Most important for the Sanhaja was control of trans-Saharan trade. But in the first half of the eleventh century this control was threatened in the south by Ghana, and in the north by their old rival, the Zanata nomads. The situation was further complicated by the fact that the unstable Sanhaja polity which had been created earlier had by now ceased to exist. The Jaddala, Lamtuna and Masufa acted independently. In these conditions the work and preaching of Ibn Yasin opened the way to the surmounting of internal conflicts, and also of economic and external political difficulties (Abun-Nasr, 1971:92; Levtzion, 1971:127-9). Prospects of robbery, the seizure of booty and settling the score with opponents further promoted proselytizing ardour. Although the Almoravid leaders were deeply religious, the movement as a whole from the outset was like a large-scale predatory expedition (Hopkins, 1958:28). Indeed, the nomads were looking in the plains of Morocco for a more prosperous life than they had known in the sands of the Sahara (Abun-Nasr, 1971:92). However, as the Almoravids conquered more and more sedentary countries, they came up against the perennial question of what to do next. Relations between Abu Bakr and Ibn Tashfin, although peacefully resolved, are very illuminating in this respect. Abu Bakr continued to lead a nomadic life, not wishing to leave the Sahara; but Ibn Tashfin, for all his piety, was already oriented towards the countries they had conquered. Whether consciously or unconsciously, Abu Bakr was an advocate of the first tendency in the development of Almoravid state, and Ibn Tashfin an advocate of the second. Although the final break between the Saharan and Maghrib parts of the empire did not take place before the death of Abu Bakr, it was already very much in evidence before his death. The short-term conquest of Ghana was made very much in the style of a nomadic state of the first type. But after the death of Abu Bakr the Sanhaja political unit in the Sahara really did disintegrate and the Almoravid movement there quickly subsided (Levt­ zion, 1971:130). Meanwhile the Saharan Sanhaja returned to their customary infighting, which earlier the leaders of the movement had managed with great difficulty to contain. But neither could the second tendency be at all fully realized in the state of Ibn Tashfin. The centre of the Empire was moved to sedentary areas, but in the long-term this weakened rather than strengthened the Empire. Ibn Khaldun stated directly that the Almoravids' source of reinforcements was to be found in the desert; but the more they were drawn into the affairs of the Maghrib and Spain, the weaker their ties with the Sahara and the nomads still living there became, and the less reinforcements they received from the Sahara (Levtzion, 1977:337, 338). At the same time in both the Maghrib and Spain different strata of the 286

The Near East population regarded the Almoravids as foreign conquerors. No firm union or merging of the conquerors and the local ruling classes took place. The situation was further complicated by the fact that the Almoravids had subjugated not only the sedentary population, but also the nomads. They preferred to settle the old conflict between the Sanhaja and Zanata by force, although there was little to be gained from this. In areas inhabited by nomads taxes could only be collected by military force, thus they could remain unpaid for years (Hopkins, 1958:54). But by allowing other nomads no part in the benefits of their conquest, the Almoravids were pulling the rug out from under their own feet (Le Tourneau, 1969:13). The Sanhaja regarded themselves as the ruling class, and indeed they did become such. They even forbade their subjects to wear the litham (veil), which they preserved as their own distinguishing mark. They received various benefits and privileges from the dynasty; they provided military commanders who were at the same time the governors of the provinces. It was under the Almoravids that the practice of giving land-rewards for service was first introduced (Hopkins, 1958:83). Admittedly, the Almoravids could not dispense with their educated Andalusian secretaries. In fact, the role of the Andalusians in their state apparatus resembles the role of the Persian bureaucracy in the Saljuq Empire, although the influence of the former was more limited. But even amongst the Sanhaja themselves the distribution of privileges was uneven. The Lamtuna were the most privileged tribe and formed the nucleus of the army and the new aristocracy. In the social and political respect the Almoravid Empire represented the hegemony of one tribal group of nomadic origin over a large sedentary and nomadic population (Abun-Nasr, 1971:101). But this social base was too narrow. For all their religious fanaticism and wars against Christians, the Almoravids were the first in North Africa to use troops made up of Christian mercenaries (Hopkins, 1958:54; Levtzion, 1977:334). But, of course, this was not a real solution to the problem. The son and successor of Ibn Tashfin, Ali, had not grown up in the desert, but in an urban Andalusian milieu; however, the social basis of his state became no wider on this account. 'In brief, behind the beautiful facade of the Almoravid empire lay a society in transition and the different parts, assembled by the decree of Yusuf ibn Tashfin, did not fit together very well' (Le Tourneau, 1969:14-15). The Almohad Empire was created not by the nomads of the Sahara, but by mountain agriculturalists, the Masmuda. But nomads did play an ever increasing role in the Empire, while at the same time they were an ill controlled and ill managed force within it. By the middle of the twelfth 12

" According to Gcllner (personal communication) the whole Sanhaja/Masmuda/Zanata theory had been Utently exaggerated by French historiography.

287

Nomads

and the state

century the Banu Hilal had already established themselves in Ifriqiya and partly also in the Central Maghrib, and the Banu Ma'qil had moved along the borders of the Sahara to the Far Maghrib. The Zanata were usually the allies of the Arab nomads. All that the government tried to do was to set different groups of nomads against one another (Abun-Nasr, 1971:110, 112). In order to weaken the power of the Arab tribes in Ifriqiya and the Central Maghrib, Abd-alMumin moved some of them to Morocco and gave them lands between Rabat and Casablanca, hoping thereby to create a basis of support for his dynasty (Marcais, 1913:187; de Planhol, 1968:140; Le Tourneau, 1969:66). As a result the Arabs and Arabized Zanata quickly became the major military force in the state. They received pastures and arable lands as appanages and not only were they exempted from paying taxes, they also exacted them from the sedentary population (Margais, 1913:250; Julien, 1956a: 111-12; Levtzion, 1977:343) and did not miss any opportunities for direct robbery in the process. Their loyalty to the dynasty was entirely dependent on what they considered to be to their own best advantage. In the last period of the Almohad state their influence was greatly increased because they actively participated in dynastic quarrels and supported certain claimants against other ones. Whether or not the disintegration of sedentary life was caused by nomads or, on the contrary, nomads only made use of the circumstances of the decay which had begun earlier (de Planhol, 1968:141; Le Tourneau, 1969:103^4), under the Almohads nomads securely occupied over a long period the privileged positions in the states of North Africa, irrespective of the circumstances in which these states had emerged. Under the Abdalwadids, Marinids and Hafsids the role of nomads further increased. The first two dynasties themselves were descended from aristocratic nomadic clans and, therefore, the nomadic aristocracy laid claim to an exclusive position in these states. But even amongst the Hafsids (Brunschvig, 1940; Brunschvig, 1947) Bedouin sheikhs were placed on the same footing as the highest stratum of the aristocracy, the descendants of the Almohads. The Hafsid rulers did not try to interfere in the internal affairs of the nomadic tribes. On the other hand it was in the Hafsid state that, from the end of the thirteenth century, lands were handed out to nomads as iqta on a very large scale (Julien, 1956a: 140). In the fourteenth century entire provinces had already been turned into iqta. The attempts of individual sultans to oppose this process were unsuccessful. Rivalry between the different nomadic polities made their situation even worse. In effect, sedentary life in the inner areas of the state decayed and the sedentary inhabitants of the plains began to retreat into towns (de Planhol, 1968:152-4). It was only by the sea that towns flourished 288

The Near

East

and then they were oriented towards external trade and out of touch with their own hinterland. After the Zanata tribe of the Banu Marin won its first victories over the Almohads at the beginning of the thirteenth century, it no longer returned to the desert but embarked on a campaign to take Morocco for, in common with many other nomads, it was attracted by the rich pastures and prospects of booty there. Before he had even conquered the country the Banu Marin chief, Abu Yahya, had divided it up into areas which he designated to individual subdivisions of the tribe. Later on the Marinids had to rely on the support of individual Zanata and Arab tribes while at the same time they were fighting with other nomadic tribes. The Marinid Abu-l-Hasan was able to establish power for a short period over all of the Maghrib; he tried to abolish the right of the Arab tribes to collect taxes from agriculturalists and even attempted to control their pastoral migrations. However, he was defeated by nomads in 1348 at the Battle of Qairawan (Abun-Nasr, 1971:129). The Abdalwadid state (Julien, 1956a) was also founded by Zanata nomads. Although the founder of the state, Yaghmurasan ibn Zayyan, who settled in Tlemcen, gave lands to his Arab subjects, the first Abdalwadids stressed their origin in all manner of ways. They remained sheikhs in their own tribe, embarked on campaigns at the head of the tribe and made a demonstration of consulting other sheikhs. Only Abu Hamuw was able, at the beginning of the fourteenth century, to complete the transition from nomadic sheikh to sultan of a sedentary population. Yaghmurasan had already tried to make the nomads leave his state. However, his failure to do so led only to new incursions and to his inviting other tribes into the state where they quickly acquired power and influence. According to Ibn Khaldun, in 1380 the real power of the Abdalwadid sultans lay in the coastal territories. All their other lands had been distributed as iqta. In the specific conditions of North Africa both paths open to nomadic statehood were essentially dead ends. Immediately after conquering a sedentary society nomads no longer felt any need for centralized power. This was because in the sedentary world there was no force which could oppose them, so the need for a state of the first type was no longer operative. Thus, if the leaders of a conquest were to avoid becoming once again ordinary nomadic sheikhs, they had to move into the towns and become rulers of sedentary populations, that is they had to make a step towards the second tendency. They tried to rule sedentary populations, but could not dispense with their main support in the state, which came from their fellow tribesmen or the tribes in their guishes. However, nomads wanted neither to settle, nor to become fully integrated in the sociopolitical system of the new state, and amongst the 289

Nomads

and the state

sedentary strata in the state there were no forces which could force them to do so. 'No definite solution was ever found for either of these problems. Successive Moroccan dynasties used a number of different principles of recruiting their military support; their tribes of origin; foreign (e.g. Christian) mercenaries; imported Arab tribes that had been defeated in expeditions abroad; trained Negro slave armies; or a combination of privileged tribes and standing army' (Gellner, 1969:4). But all these attempts were failures. Only the Ottoman Turks were more successful than the others at perpetuating their supremacy, and they established themselves relatively securely in Tunis and Algeria. The question as to why they were able to do this deserves separate study. I shall only point out here, without any claim to be replying fully to the question, that the strength of the Ottoman Empire was incomparably greater than that of any other medieval state in North Africa. Thus, it happened that while nomadic dynasties represented the interests of sedentaries, at the same time they had to rely on nomads and make it possible for the latter to exploit agriculturalists and townsmen. Whether or not they wanted to do this and despite the fact that they had lost their connections with nomadic life and, essentially, their interests had become those of a sedentary state, the nomadic dynasties could not definitively turn their states into states of the second type. It was even more difficult for them to reconcile the different interests within these states. It was only in the late Middle Ages that individual rulers were able temporarily to limit the destructive role played by nomads in their states. The Moroccan Mulay Ismail was somewhat similar to Shah Abbas I in this respect. But the activities and successes of the former were only a short-term historical episode. Thus, social relations amongst North African nomads were fairly contradictory. With regard to the state as a whole the guish tribes were a privileged estate in a class society, and the nomadic aristocracy was one of the ruling classes. But with regard to their fellow tribesmen this aristocracy was only a leading stratum. Looking at these tribes from within, one might come to the conclusion that although socially differentiated they were still pre-class societies or sub-societies when in one way or another they were involved in a state. However, since the incorporation of different nomads within different states was not taking place all the time, considerable oscillations in the social processes amongst North African nomads are to be observed throughout their history in the Middle Ages. East Africa The problems linked to the emergence of nomadic or, to be more precise, pastoral statehood in this region essentially concern the history of the states 290

East

Africa

of the Interlacustrine region. In the past such general theories as the 'Hamitic theory' and 'conquest theory', which nowadays are no longer reputable, have been founded on the existence of these states. In this book I have already discussed the conquest theory of the origin of the state. The Hamitic theory, in the form in which it is set out by Speke (1864) and his followers, can now be considered to have been disproved (see, for example, Macgaffy, 1966; Sanders, 1969). However this does nothing to clarify the problem of the emergence of the Interlacustrine states. On the contrary rather, never has the controversy about these circumstances been so great as in the past few years. I make no claim to solve the problem in all its complexity here. At present there are insufficient data for this to be possible. Instead I try to group together known facts and hypotheses which, in my view, correspond to these facts and on this basis I try to come to certain preliminary conclusions. The sociopolitical organization and even the economy of the states and chiefdoms of the Interlacustrine region on the eve of the colonial period were not nearly as uniform as they were thought to be not so very long ago. The fact that there were certain common political and ideological institutions cannot hide essential differences (Beattie, 1971:245ff.); nor does it mean that the states of the Interlacustrine region were at an identical stage of development, even if they have been defined by different scholars as feudal or early class, 'Asian' or 'African', despotic or bureaucratic. Comparisons only need to be made between Buganda and Rwanda, and Bunyoro and Ankole for the differences to become clear. The historical development of each state in the Interlacustrine region has its own individual distinguishing features. Nor is it inconceivable that the emergence of each of these states had local peculiarities. With the exception of Buganda, the states of the Interlacustrine region were ethnically heterogeneous. In the majority the population may be divided up into groups which may conditionally be called ethno-social strata: Tutsi (Hima, Huma), Hutu (Iru) and in some places the Twa. The physical differences between these strata are so noticeable that they can hardly be explained otherwise than as differences of origins. Those who are inclined to regard them as the consequence of social differences resulting from economic specialization and differences in dietary systems (Posnansky, 1966a:6) follow Lysenko or, at best, Lamarck, but certainly not contemporary genetic theories (cf. Hiernaux, 1966). On the contrary physical anthropological data and Tutsi links with pastoralism, also the species of large stock, the sanga, which they herd, are fairly reliable indications that they were of northern descent. Evidently the appearance of their ancestors in East Africa was connected to the numerous migrations there of pastoral tribes. It need not be supposed that they all came with one migration; indeed, the ancestors of the Tutsi may well have 291

Nomads

and the state

come to the Interlacustrine region at different times and from different places (Mair, 1962:129; d'Hertefeld, 1962:17; d'Hertefeld, 1965:406; Lemarchaund, 1966:599; Posnansky, 1967:645; Oliver, 1977:626ff.). But, of course, the final formation of the Tutsi proper occurred in East Africa and with the Bantu-speaking ancestors of the Hutu as one of their components. This and the ensuing ethnic mixing explain the language assimilation of the Tutsi and the appearance of intermediary anthropological types (d'Herte­ feld, 1962:17-18). There is no foundation for the suggestion (see, for example, Oberg, 1940:121; Oliver, 1961:58; Dunbar, 1965:24) that pastoralists brought statehood to the Interlacustrine region in an already completed form. However, it should not necessarily be supposed that there was little social differentiation amongst them as was the case amongst other East African pastoralists such as, for example, the Karimojong or the Turkana. If the local tradition that the Bacwezi were Kulturtrageren (Nyakatura, 1973:17ff.), in contrast to the tradition about the Batembuzi (cf., however, D. W. Cohen, 1972:80ff.), has any real foundation, and it would appear that the majority of specialists think it does (cf., however, Wrigley, 1958:16; Henige, 1974:44-5), then it should be said that in the fourteenth to the fifteenth centuries A . D . Uganda was inhabited by a group of pastoralists at the head of which there stood the Bacwezi lineage or clan (Dunbar, 1965:2, 24; cf. Steinhart, 1978:132). Another possible variant, which to me seems less convincing, is that Bacwezi is a name for both pastoralists and their leaders. However, local legends differentiate the Bacwezi and Bahima and stress that the former were not numerous. There are also other arguments for not putting them together (Posnansky, 1967:642; Nyakatura, 1973:249; cf. Dunbar, 1965:21f., 45). Therefore, as a whole, the relationship of the Bacwezi to the Bahima remains a complex and unsolved problem. In all circumstances archaeological data (about Bigo and other sites with their 'great earthworks', the construction of which required the concentra­ tion and organization of a large work force - see Posnansky, 1966a) corroborates the view that Bacwezi society was already familiar with institutionalized social inequality. But if there was social differentiation amongst pastoralists either before they appeared in Uganda, or during the period in which they were establishing themselves there, there must have been certain reasons for this. They could have been connected with their pre-history or with the circumstances of their migration and settlement in Uganda or, finally, with the history of the period when they lived there. The disappearance of the Bacwezi from Uganda is probably best 1 3

1 3

All dates connected with the period in the history of the Interlacustrine region which is being examined here should be regarded as tentative.

292

East

Africa

explained by their further migrations, undertaken for reasons we do not fully understand and which perhaps were connected with pressure from new groups of pastoralists (the Luo?) from the north at about the end of the fifteenth century (Posnansky, 1967:634; Ogot, 1967:46). This pressure was aggravated by natural disasters (Dunbar, 1965:24). It would seem that the Bacwezi still remained if not real nomads, then at least mobile pastoralists. However, the legendary memory which they left behind them in the Interlacustrine region and the striving of various dynasties in several states there to link their descent, whether or not they had the grounds to do so, with the Bacwezi (Posnansky, 1966a:5; Beattie, 1971:57; Oliver, 1977:636) demand some kind of explanation. It is possible that in the Interlacustrine region under the Bacwezi, apart from economic links, some sort of sociopolitical ties between pastoralists and agriculturalists were secured, and if the latter were not the first to be established, at least they were the first to be clearly defined so that for a period the Bacwezi were able to rule both pastoral and agricultural groups. The legendary tradition of the Kitara 'Empire' derives from this. The only point in this problem on which the majority of scholars are able to agree is that the interaction between pastoralists and agriculturalists played an important role in the emergence of the 'Empire'. Admittedly, it is sometimes maintained that the agriculturalists of the Interlacustrine region had their own statehood in the period preceding the pastoral migrations (Wrigley, 1958:16,17; H. S. Lewis, 1966a:405; Katoke, 1971; cf. Murdock, 1959:350); however, there are no serious grounds for this view (Steinhart, 1978:134). The most that can be allowed is that there were small chiefdoms. The Interlacustrine states emerged later on, after the pastoral migrations and, moreover, not immediately after. The appearance of the Nilotic Luo in the Interlacustrine region and the changeover from the Bacwezi to the Bito dynasty (in Bunyoro, as the Kitara became known after the Luo had appeared) did not mean a complete return to the situation as it had existed before the Bacwezi. On the contrary, in many ways continuity of development was preserved. In the south states and chiefdoms whose dynasties were linked to the Hima were able to emerge, perhaps in part as a reaction to the Luo migration and the Bito's establishment of power. Be that as it may, the migrations of pastoralists, which could have been led by the Bacwezi, and the later pastoral migrations resulted in the establishing of economic interdependence between pastoralists and agriculturalists on the territory of the Interlacustrine region. From a purely economic point of view agriculturalists obviously were just as interested in procuring cattle and (he products of a pastoral economy as pastoralists were in procuring agricultural produce, and possibly handicraft goods as well (on the latter sec: Elam, 1974:160, 167). In certain areas cattle were also 29.3

Nomads

and the state

important for fertilizing the fields (Oliver, 1977:640). But everywhere the ownership of cattle acquired a prestigious and corresponding social significance. Service (1975:120-3) stresses the contractual nature of the balanced relations of political co-operation established in Ankole which were to some extent also beneficial for the Iru. However, the division of labour between pastoralists and agriculturalists was not established on a purely economic exchange basis; by necessity it also had to involve the establishment of certain social and political ties. The emerging single sociopolitical organization could be created by both peaceful and unpeaceful means (Lemarchand, 1966:598). In some cases, as in Buganda, the balance of power favoured agriculturalists and the status of the Hima pastoralists was very low. In others, pastoralists had the advantage. Moreover, in various states this balance did not always remain constant, and the ruling dynasties of such states could be of a descent different from that of the majority of the ruling class. It is significant, however, that the position of pastoralists was weakest of all in those areas where there was less good pasture and more land suitable for cultivation (Mair, 1962:30). Thus, Buganda is ill-suited to the development of a pastoral economy (Kottak, 1972:373). The high rainfall in this area favours tropical agriculture; at the same time 32% of the land in Uganda is infested with the tsetse fly (Baker, 1963:20). On the other hand, in Ankole most of the country was given over to pasture (Elam, 1973:162ff.). Wherever pastoralists finally turned out to be stronger than agricultural­ ists, the emerging single sociopolitical system was founded on a division of labour between them from which pastoralists could derive the most benefit. To a greater or lesser degree this involved a monopoly of the ownership of cattle; furthermore, the social division of labour became compulsory, and redistributive mechanisms strengthened the position of the dynasties and ruling elites with which they were connected (Oberg, 1940:130; Maquet, 1961:164-5; d'Hertefeld, 1965:423ff.; Lemarchand, 1966:608; Elam, 1973:161, 171). The main tendencies in social stratification coincided with economic specialization and ethnic differences; this resulted in a more or less defined tendency towards the forming of castes. Thus, it was in East Africa that the third tendency in the emergence of nomadic (pastoral) statehood became the predominant one. Not all pastoralists, of course, were included in the ranks of the ruling class, nor were all agriculturalists denied access to it. Limited social mobility always existed. It plays a part in every society. But in the states of the Interlacustrine region it is usually seen to occur within the framework of an 14

14

Amongst the Iru of Ankole there existed a legend that they had once owned cattle, but thai it had been taken away from them by the Mima (OhcrR, 1940:130; cf. Steinharl, I W M 4 2 ) .

294

Conclusions established social-ethnic cliche based on the opposition between the Tutsi and the Hutu (d'Hertefeld, 1971:56-7; Steinhart, 1978:145). As I have already pointed out, none of this is any more than a general hypothetical outline. The formation and further development of each of the states of the Interlacustrine region had their own individual characteristics. In some of them the divide between the Tutsi and the Hutu can be observed more clearly than in others, and in some it was practically non-existent (Mair, 1962:134-7;Beattie, 1971:139; Steinhart, 1978:147). InBunyoro.for example, after the sharp decrease in the number of cattle in the nineteenth century the division between the Hima and Iru was losing much of its social meaning. Anybody who owned large stock became a Hima (Beattie, 1971:24). However, it must be pointed out that an essential transformation took place amongst pastoralists, even in those states in which they occupied the dominant position. Not only dynasties and their immediate circle, but large numbers of nomads as well actually went over to the sedentary way of life. Thus, in a very distinctive form, the second tendency manifested itself also in the history of the states of the Interlacustrine region. Consequently, the emergence and transformation of these states as a whole underline the general patterns of the emergence and development of statehood amongst nomads and, more widely, extensive and mobile pastoralists. The specific character of its interaction with the outside world or specific integration with it (or within it) was the most important precondition and feature of nomadic statehood in East Africa. Conclusions The materials examined in this chapter show that only in individual, very rare cases has a nomadic state emerged without conquering and subjugating a sedentary population. Such states were little more than short-lived historical episodes. They emerged as the result of very specific problems facing nomadic society, usually on the eve of conquest and for the sake of that conquest, but it was only actual conquest itself which rendered these states stable. Exactly as the chiefdoms which preceded them may be called dispositional, so also may such states be called dispositional. A little time ago Claessen and Skalnik (1978a:627) wrote: 'It is not very clear whether conquest leads to the formation of the early state or, conversely, the foundation of the early state leads to conquest.' Instances of both are to be observed amongst nomads, but the first is considerably more common. The role played by conquest and subjugation, and violence in general in the emergence of nomadic states is so marked that, in the 'conflict-integration' controversy, the former is plainly the winner. But 295

Nomads

and the state

'conflict' in the case of nomadic societies was in the first place directed outside those societies. In the evolutionary and even in the structural respect dispositional states are very similar to 'the inchoate early states' as defined by Claessen and Skalnik (1978a:629-33; 1978b:641). The differences between them are outlined in the extent of their stability and in the course of their further evolution. In so far as a nomadic state emerges only if it has specific relations with the outside world and for certain kinds of such relations, it cannot exist for any prolonged period just because of the internal development of the nomadic societies. To exist and maintain stability over a prolonged period a nomadic state must incorporate within itself a part of this outside world in the form of its sedentary population or, either directly or indirectly, it must subjugate that sedentary population. However, for this very reason all such states are not completely nomadic states in the strict sense. Dispositional states excepted, all other nomadic states may only conditionally be referred to as nomadic. They are nomadic only in the sense that they were founded by nomads and the ruling position in them was occupied by nomads. But even to say that nomads in general were supreme in these states is inappropriate. It should always be ascertained exactly who was supreme, over whom this was and in what form it was manifested. The first social demarcation to arise in nomadic states tends to be between nomads and the sedentary population. Another characteristic feature of nomadic states is their considerable heterogeneity. If not in all, then in the majority of such states there are several ruling and dependent strata and classes, and frequently intermediary strata as well, 'middle classes' of a sort. As a rule the strata and classes in nomadic states form two sub-systems (or two sub-societies) linked together mainly by political ties. Cases of a deeper and more organic synthesis are known. However, firstly, they are usually secondary and derived from those political relations mentioned above and, secondly, they rarely develop to the full so that the two sub-systems merge into one. The social character of a sedentary sub-society was determined mainly by the relations which had existed within it before its subjugation or conquest by nomads. If, after the event, sub-societies were only made politically dependent they usually then underwent few if any serious changes, particularly if this dependence was one of simple vassalage. Even where conquest involved the destruction of a sedentary state and the incorporation of its population into a new nomadic one, the basic changes which took place were in the political sphere. Some changes did occur in the social composition of the sedentary population. But basically they affected the ruling strata and classes, rather than the dependent ones. Sometimes certain modifications in the social and 296

Conclusions economic system of a sedentary sub-society also took place, but as a rule they were not so significant as to bring about a qualitative transformation. It was actually in a nomadic sub-society in a nomadic state that more fundamental changes could occur. Firstly such a sub-society for the first time became a class society, not always completely so, but always in a fairly distinctive fashion. The most original side to the process of class formation in nomadic sub-societies in nomadic states consists, in my view, in the fact that the different strata and groups within them did not all become classes simultaneously, nor did they do so in relation to one another. The nomadic aristocracy was usually the first to become a class, although not initially in relation to ordinary nomads, but to the subjugated and conquered sedentary population. At the same time for a more or less prolonged period they were merely the leading estate of other ordinary nomads. Often the processes of social differentiation in a nomadic sub-society went no further than this. In other cases the nomadic aristocracy became a class in relation to ordinary nomads. However, even where this happened the class nature of both sub-societies remained different. Since the time of Vladimirtsov (1934) the problem of so-called 'nomadic feudalism' has been constantly discussed. For many decades now it has been debated with particular earnestness in Soviet anthropology and history (for details on this point see Khazanov, 1975; Markov, 1976; Pershits, 1976; see also pp. 123-4 of this book). However, as a rule it is not made clear if it is dispositional states of nomads which are relatively rare and unstable polities, or nomadic sub-societies in nomadic states in the broad sense of the word, which are being discussed. But, in my view, neither the first nor the second can be characterized as feudal. As Lattimore (1979:488) has rightly pointed out: 'International terms like "feudal" emphasize similarities, but often conceal differences.' Of course, it could then be argued that the question really depends on what exactly is meant by feudalism, whether it is primarily a social-economic or a political phenomenon. However, those characteristics of feudalism which both above-mentioned schools of thought believe to be the principal ones do not exist amongst nomads. The foremost Soviet medievalist Gurevich (1970:7) has defined feudalism in the following way: 'Feudal formation is characterized by the contradiction between large-scale ownership of land and the small-scale production of peasants, and by non-economic compulsion, the necessity for which derives from this basic contradiction; since the peasant conducts a self-sufficient household, the appropriation of his surplus produce is only possible through the application of force in one or another form. Linked to this system of productive relations are such characteristics of feudalism as the conditional nature of feudal land-ownership and the hierarchy of its structure, as well as the hierarchy of the ruling class. Such an understanding of feudalism gives us 297

Nomads

and the state

a firm basis on which to study specific problems of the history of the Middle Ages.' However, further on in his book Gurevich shows that his definition is no more than an incomplete theoretical model. On the one hand, this model includes features which are characteristic not only of feudalism and, on the other hand, it is far from the case that it is always and in every way applicable even to the countries of medieval Western Europe. Gurevich came to these conclusions himself. I should now like to extend the majority of them to another definition of feudalism made by the great historian Marc Bloch (1974, n:446): 'A subject peasantry; widespread use of the service tenement (i.e. fief) instead of salary, which was out of the question; the supremacy of a class of specialized warriors; ties of obedience and protection which bind man to man and, within the warrior class, assume the distinctive form called vassalage - leading inevitably to disorder; and in the midst of all this, the survival of other forms of association, family and State But, most importantly, so far as nomads are concerned, neither definition is acceptable. In this respect one of the most recent attempts to demonstrate the feudal nature of the Mongol state, made by Munkuev, is indicative. Having pointed out that the idea of a centralized monarchy was alien to Jenghiz Khan and that for this reason his state was founded on the ulus system, Munkuev concludes: 'Therefore, in Mongolia from the outset feudalism in the classical, political [sic] sense was established (here I am not talking about the social-economic relations mentioned above) with its characteristic vassalage and with other institutions' (1977:393). I have already written about the very specific and relative nature of Mongol vassalage (see pp. 158,159-60). Absence of a centralized monarchy is clearly also insufficient evidence with which to identify the state as a feudal one. As for social and economic relations, Munkuev (1977:387) in the abovementioned work refers to three institutions: otogu-boyol, nokur and domestic slavery. None of them can be called feudal (see p. 159-60 of this book). In this case Munkuev's preconceived assessment is clearly contradicted by the facts. Since I adhere to the narrow regional and temporal view of feudalism as primarily a European phenomenon of the Middle Ages, I am also unable to call sedentary sub-societies in nomadic states feudal. However, differentiation should be made between feudalism as an entire political, social and economic system and its separate elements and institutions (vassalage, for example). Some of the latter are considerably more widespread than feudalism. Even more often they only externally coincide with feudal ones (cf. Service, 1975:82). Nevertheless, with insufficient justification a society is sometimes called feudal because such institutions exist within it. 298

Conclusions Whenever a nomadic society (or sub-society) attains that level of social differentiation which can be characterized as class becomes in my view an early class society. Moreover, in my view it will remain such as long as it is nomadic and politically independent. It can also revert once again to becoming a pre-class society. On the other hand, even when a nomadic society is subdued by a sedentary one, more developed class relations pene­ trate only with difficulty and from outside. By early class society I mean a society: (a) which in the evolutionary scale occupies the position between a pre-class (primitive) and developed class society; (b) in which there is no private ownership of key resources or in which this is not the main criterion of social differentiation; (c) in which the different relations of the basic strata, estates and classes to production and distribution are not directly linked to relations about the ownership of key resources; (d) in which the major social differences are linked to political supremacy and dependence, which in the majority of cases are determined by the different social relations to institutions of government and management in the society; (e) in which there exists a regular tax system in kind and/or where the dependent strata and classes carry out other obligations to the state and/or ruling strata and classes. Returning now to the three types of nomadic statehood singled out earlier, it should be made clear that they refer to nomadic states in the broad sense of the word, not to dispositional states of nomads. Once again I should like to emphasize that these types are first and foremost a conditional and simplified theoretical model. In practice all we can really talk about is a more or less intelligible predominance of a corresponding tendency and degree of realization. Social relations in nomadic states of the first type are determined by three main factors: (a) the relationship of the nomads as a whole to the conquered or subjugated sedentary population or sedentary state; (b) the relations between the nomadic aristocracy and the sedentary population or sedentary state; (c) the relations between the nomadic aristocracy and the ordinary nomads. Sometimes a fourth feature is added to these three - the relations between the various nomadic units, polities, tribes, ethnic groups, etc. In states of the first type relationships of dominance and submission or dependence established between nomads and, especially, their aristocracy and the sedentary population as a whole are the principal ones. Indeed, the very existence of such a state depends upon these relationships. Correspon­ dingly, external forms of dependence and exploitation predominate over internal ones. The stronger this predominance, the more completely does this tendency which is characteristic of states of the first type manifest itself. In states of the first type the nomadic and sedentary societies do not even always form separate sub-societies which are socially or sociopolitically and, frequently, geographically divided. Wherever a sedentary state continues to 299

Nomads

and the state

exist and its dependence on the nomadic state is only of the vassal-tribute type, nomads, on the one hand, and agriculturalists and townsmen, on the other, generally continue to live in two separate societies. In this case the relations between the nomadic aristocracy and ordinary nomads may acquire an early-class form, but frequently they remain pre-class. With regard to the nomadic society or sub-society proper the aristocracy frequently acts only as the leading estate, certainly not as the ruling class. Ordinary nomads continue to be the main military and social support of the nomadic aristocracy, and redistributive mechanisms continue to function in such states amongst nomads. The situation may undergo no fundamental changes if in all the abovementioned circumstances nomads and the sedentary population are integrated within a single state. In this state there exist at least two ruling classes: the new nomadic and old sedentary one (or sedentary ones), and the problem of the relations between them is a very acute one. The situation of the non-privileged sedentary populations in such states usually worsens, and their exploitation usually increases; but the situation of ordinary nomads remains basically the same. In states of the second type nomads and the sedentary population can also form two separate sub-societies, but only socially, not politically or geographically. However, integrative processes are manifested consider­ ably more strongly in these than they are in states of the first type. The tendency to do away with the boundaries between the two sub-societies even exists in the social sphere, but in practice it is nowhere near always fully realized. To some extent political synthesis is always supplemented by social synthesis, but the extent depends on the specific situation. The main factors determining the social relations in states of the second type are: (a) the relations of the ruling classes to the peasantry as a basically dependent and exploitable class; (b) the relations between the ruling classes themselves, particularly the old and new, the sedentary and the nomadic; (c) the relations of the ruling classes to ordinary nomads. If in states of the first type the position of the nomadic aristocracy (ruling class-leading estate) is first and foremost and predominantly a dual one, then in states of the second type the situation of ordinary nomads (subjects-fellow tribesmen), no less of an ambiguous one. However, not only are ordinary nomads not the basic class in such states, it is also rare for them to form a single estate or class within them. The more completely is the second tendency realized, the more differentiated the nomadic sub-society becomes. To a greater extent than in states of the first type the nomads in the second type divide up into the privileged, less privileged and non-privileged, depending on their ties with the dynasty and the nomadic aristocracy, their ethnic and tribal connections, etc. Usually they form several intermediary estates and strata, some of 300

Conclusions which are closer to the ruling classes, and others to the dependent ones. Thus it is not always possible to say that the government or ruling classes in such states have a specific policy for ordinary nomads. Usually their policy differs with regard to the different strata, units and groups of nomads in the different periods of a state's existence. With regard to a nomadic sub-society the integrative process in states of the second type can assume different forms. But on the whole it is marked by social unevenness and a lack of simultaneous impact. Usually the dynasty and its immediate entourage who head the state are the first to be affected by them, then all or part of the nomadic aristocracy which becomes the ruling class of the sedentary population, or one of its ruling classes. As the integrative processes develop, the nomadic aristocracy and, particularly, the dynasty have to decide whether to direct their attention at the ruling position in the state and to combine their interests with those of the state as a whole and with the sedentary sub-society, or whether in the first place to try to preserve their former position in the nomadic sub-society and in doing so sometimes act against the interests of the state. The dilemma is never an easy one and not always is a unanimous decision taken on it, even by the various strata and groups of the nomadic aristocracy within one state. As a result new conflicts frequently emerge in states of the second type, between the dynasty of nomadic origin and. its supporters, on the one hand, and the traditional nomadic aristocracy, on the other, between the dynasty and ordinary nomads, etc. States of the second type do not always remain nomadic even in the broad sense of the word. One widespread way in which they are transformed is by turning into a sedentary state in which some nomads gradually become sedentary, while others, from the distinct, politically dominant sub-society, gradually turn into the backward social and sometimes also ethnic minority which is encapsulated in the more developed socioeconomic and political system of a sedentary state. The word 'nomadic' can only be applied very conditionally to states of the third type. With regard to the third type, far more than to the first and second types, it is more precise to talk about a corresponding tendency rather than about a type in any distinct form. A full manifestation and the corresponding realization of the third tendency only occurs with the conjunction of two different processes: the emergence of a single social, economic and political system in which both nomads and agriculturalists are integrated, and the formation of social differences along lines which basically coincide with economic specialization and frequently are linked in addition to ethnic differences. These processes occur separately far more often than they do together. This is conceivably the reason why, in the emergence and transformation of nomadic statehood, the third tendency occurs less frequently than the first two and, when it 301

Nomads

and the state

occurs, it is even less frequently the dominant tendency but rather occurs in different combinations with the others. Another reason which also makes it difficult to discuss states of the third type is that the creation of a single socioeconomic system, incorporating nomads or pastoralists with the sedentary population within it, in itself gives these states a distinct mixed form. In practice that process usually leads to the politically dominant nomads ceasing to be nomads, although they may preserve their pastoral traditions as part of their cultural heritage and/or the criteria of their social distinctions. Nevertheless, nomads or extensive pastoralists usually play the leading role in the processes cited, especially in their fusion. It is, in effect, for this reason that we can take the third tendency into account when examining the problems connected with nomadic statehood.

3(12

By way of a conclusion: the outside world and nomads

The problem of the interrelations between nomads and the sedentary agricultural and urban world essentially has two sides to it, about both of which a great deal has been written. The most important research into the first side of the problem has been done by anthropologists, and I have largely followed in their footsteps. The research into the second has until now been almost completely dominated by historians; nevertheless, many related questions still remain not only unclear and debatable, but also in need of anthropological study. I shall limit myself to just a few examples. It is neither possible nor necessary to become involved at this point in the details of the complex and debatable problem of the nature of Chinese, Middle Eastern or Near Eastern societies in the Middle Ages. Was the general direction of the evolution and tempo of evolution in these societies affected by the fact that they were frequently conquered by nomads? If so, in what way? These are all themes for further research rather than problems which have already been solved. Even the question of the frequently claimed link between the nomadic conquests and the slowness of the development of the societies cited is far from simple. Slowness in comparison with what? Clearly not with Africa or pre-Colombian America: consequently, slowness in comparison with Europe. But is it right anyway for the question to be put in this way? Surely it could be formulated otherwise. What causes the quick tempo of development in European societies? Is it the absence of nomadic invasions? But, in the first place, in the late palaeolithic epoch and again in antiquity the development of Europe outstripped that of many other regions. And, secondly, absence of nomadic invasions in Europe was fairly relative. Somehow or other the Huns were able to inflict fatal damage on the Roman Empire, while the Avar raids and Magyar conquest of Pannonia in the ninth and tenth centuries reverberated in most of Europe (Dienes, 1972). Nor should what in Russian historiography has become known as 'the three centuries of the Tatar yoke' be forgotten. Finally, a quicker or slower tempo of development can be properly

By way of a

conclusion

discussed only if we are dealing with unilinear evolution, that is if Chinese and the different Muslim societies were feudal, like medieval European society. (Incidentally, European feudalism was also fairly multi-faceted.) But this problem has been discussed since the time of Voltaire and Montesquieu and is still far from its final solution. If the evolution of medieval societies in Asia and Europe (and neither should we forget ancient societies) was multilinear, where then are the criteria and measure by which we are to gauge its tempo? Are they comparable at all? At any rate it is doubtful that every nomadic conquest inevitably slowed down the evolution of the society it had conquered. The Qarakhanid and Saljuq conquests brought no great destruction or upheaval. Many wars and conquests between sedentary states were more bloody and ruinous. With regard to the social and economic sphere the problem can sometimes be looked at the other way round. Bearing in mind the fact that it was as a result of the conquests mentioned above that the institution of iqta, which had existed under the Buyids and Samanids, became much more widespread, it could be asked whether these conquests slowed down the tempo of evolution in the corresponding societies, or whether they accelerated it. I do not insist upon either view. All I wish to show is that there are no grounds for categoric assertions, but on the other hand there is a great deal of room for discussion and further research. This fully applies to all the questions and problems which are raised in my book. Here to the best of my ability I have tried to suggest solutions to these problems which at the present time seem to me to be the most acceptable ones. But this does not mean that my colleagues will see them as such, or, possibly, that I myself will regard them in the same light in the future. This is because I should like to think that I shall have the possibilities to continue to work on the inexhaustible subject of nomads and nomadism. And on this point I must end. * After August 19791 was able to introduce certain minor amplifications and additions, mainly bibliographical ones, into the manuscript.

304

Abbreviations

AA AQ CA CERM CHA CHI GR ISSJ JA AS JAH JESHO JRGS PPP SJA WA

American Anthropologist Anthropological Quarterly Current Anthropology Centre d'6tudes et de recherches marxistes. Paris The Cambridge History of Africa The Cambridge History of Iran The Geographical Review International Social Science Journal Journal of Asian and African Studies Journal of African History Journal of the Economic and Social History of the Orient Journal of the Royal Geographical Society Paleogeography, Paleoclimatology, Paleoecology Southwestern Journal of Anthropology World Archaeology

VDI Vf VLU

Vestnik drevnei istorii [Bulletin of Ancient History] Voprosy istorii [Problems of History] Vestnik Leningradskogo Universiteta [Bulletin of Leningrad Univer­ sity] Kratkie soobshcheniia Instituta arkheologii Akademii Nauk SSSR (Short Reports of the Institute of Archaeology of the Academy of Sciences of the U S S R ] Kratkie soobshcheniia Instituta vostokovedeniia Akademii Nauk SSSR [Short Reports of the Institute of Oriental Studies of the Academy of Sciences of the U S S R ] Kratkie soobshcheniia Instituta istorii materialnoi kultury Akademii Nauk SSSR [Short Reports of the Institute of the History of Material Culture of the Academy of Sciences of the USSR] Kratkie soobshcheniia Instituta etnografii Akademii Nauk SSSR [Short Reports of the Institute of Ethnography of the Academy of Sciences of the U S S R ] Leningradskii Gosudarstvennyi Universitet [Leningrad State Uni­ versity] Moskovskii Gosudarstvennyi Universitet [Moscow State University] Matcrialy i issledovaniia po arkheologii SSSR [Materials and Studies